0% found this document useful (0 votes)

214 views

Pors Hyper Linked

This document provides an introduction to and overview of the textbook "Principles of Remote Sensing". It discusses remote sensing as the acquisition of spatial data using sensors mounted on platforms, and the wide range of applications of remote sensing data. The textbook is structured to cover topics including electromagnetic energy, sensors and platforms, aerial cameras, multispectral scanners, radar, subsurface remote sensing, radiometric and geometric aspects of remote sensing.

Uploaded by

goomey

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

214 views

Pors Hyper Linked

Uploaded by

zoom

contents

index

about

Contents
1

Introduction to remote sensing

1.1
1.2
1.3

25
L. L. F. Janssen
Spatial data acquisition . . . . . . . . . . . . . . . . . . . . . . . . . 26
Application of remote sensing . . . . . . . . . . . . . . . . . . . . . 32
Structure of this textbook . . . . . . . . . . . . . . . . . . . . . . . . 43

Electromagnetic energy and remote sensing

2.1
2.2

2.3

first

Introduction . . . . . . . . . . . . . . . . .
Electromagnetic energy . . . . . . . . . . .
2.2.1 Waves and photons . . . . . . . . .
2.2.2 Sources of EM energy . . . . . . .
2.2.3 Electromagnetic spectrum . . . . .
2.2.4 Active and passive remote sensing
Energy interaction in the atmosphere . . .
2.3.1 Absorption and transmission . . .
2.3.2 Atmospheric scattering . . . . . .

last

back

exit

.
.
.
.
.
.
.
.
.

zoom

.
.
.
.
.
.
.
.
.

contents

.
.
.
.
.
.
.
.
.

49
T. Woldai
. . . 50
. . . 52
. . . 53
. . . 56
. . . 58
. . . 60
. . . 61
. . . 63
. . . 65

index

about

Contents
2.4
3

Sensors and platforms

3.1
3.2

3.3

3.4
3.5
4

Energy interactions with the Earths surface . . . . . . . . . . . . .

2.4.1 Spectral reflectance curves . . . . . . . . . . . . . . . . . . .

Introduction . . . . . . . . . . . . .
Sensors . . . . . . . . . . . . . . . .
3.2.1 Passive sensors . . . . . . .
3.2.2 Active sensors . . . . . . . .
Platforms . . . . . . . . . . . . . . .
3.3.1 Airborne remote sensing . .
3.3.2 Spaceborne remote sensing
Image data characteristics . . . . .
Data selection criteria . . . . . . . .

.
.
.
.
.
.
.
.
.

83
L. L. F. Janssen & W. H. Bakker
. . . . . . . . . . . . . . . . 84
. . . . . . . . . . . . . . . . 85
. . . . . . . . . . . . . . . . 87
. . . . . . . . . . . . . . . . 97
. . . . . . . . . . . . . . . . 101
. . . . . . . . . . . . . . . . 102
. . . . . . . . . . . . . . . . 104
. . . . . . . . . . . . . . . . 107
. . . . . . . . . . . . . . . . 109

Aerial cameras
4.1
4.2

4.3

first

Introduction . . . . . . . . . . . . . . . . . . . . . . .
Aerial camera . . . . . . . . . . . . . . . . . . . . . .
4.2.1 Lens cone . . . . . . . . . . . . . . . . . . . .
4.2.2 Film magazine and auxiliary data . . . . . .
4.2.3 Camera mounting . . . . . . . . . . . . . . .
Spectral and radiometric characteristics . . . . . . .
4.3.1 General sensitivity . . . . . . . . . . . . . . .
4.3.2 Spectral sensitivity . . . . . . . . . . . . . . .
4.3.3 Monochrome photography . . . . . . . . . .
4.3.4 True colour and colour infrared photography

last

back

exit

70
72

zoom

.
.
.
.
.
.
.
.
.
.

contents

.
.
.
.
.
.
.
.
.
.

118
J. A. Horn
. . . . 119
. . . . 122
. . . . 123
. . . . 125
. . . . 127
. . . . 128
. . . . 130
. . . . 131
. . . . 132
. . . . 134

index

about

Contents

4.4

4.5
4.6
5

4.3.5 Scanning . . . . . . . . . .
Spatial characteristics . . . . . . .
4.4.1 Scale . . . . . . . . . . . .
4.4.2 Spatial resolution . . . . .
Aerial photography missions . .
Advances in aerial photography

.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

154
W. H. Bakker
. . . . . 155
. . . . . 156
. . . . . 157
. . . . . 159
. . . . . 160
. . . . . 162
. . . . . 163
. . . . . 164
. . . . . 165
. . . . . 166
. . . . . 168
. . . . . 171
. . . . . 172
. . . . . 173
. . . . . 175
. . . . . 177

Multispectral scanners
5.1
5.2

5.3

5.4

first

Introduction . . . . . . . . . . . . . . . . . . . . . . . .
Whiskbroom scanner . . . . . . . . . . . . . . . . . . .
5.2.1 Spectral characteristics . . . . . . . . . . . . . .
5.2.2 Geometric characteristics . . . . . . . . . . . .
Pushbroom scanner . . . . . . . . . . . . . . . . . . . .
5.3.1 Spectral characteristics . . . . . . . . . . . . . .
5.3.2 Geometric characteristics . . . . . . . . . . . .
Some operational spaceborne multispectral scanners .
5.4.1 Meteosat-5 . . . . . . . . . . . . . . . . . . . . .
5.4.2 NOAA-15 . . . . . . . . . . . . . . . . . . . . .
5.4.3 Landsat-7 . . . . . . . . . . . . . . . . . . . . .
5.4.4 SPOT-4 . . . . . . . . . . . . . . . . . . . . . . .
5.4.5 IRS-1D . . . . . . . . . . . . . . . . . . . . . . .
5.4.6 IKONOS . . . . . . . . . . . . . . . . . . . . . .
5.4.7 Terra . . . . . . . . . . . . . . . . . . . . . . . .
5.4.8 EO-1 . . . . . . . . . . . . . . . . . . . . . . . .

last

back

exit

zoom

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

contents

.
.
.
.
.
.

index

.
.
.
.
.
.

138
139
140
142
143
147

about

Contents
6

RADAR
6.1
6.2
6.3

6.4

6.5

6.6
7

What is radar? . . . . . . . . . . . . . . . . . . . .
Principles of imaging radar . . . . . . . . . . . .
Geometric properties of radar . . . . . . . . . . .
6.3.1 Radar viewing geometry . . . . . . . . . .
6.3.2 Spatial resolution . . . . . . . . . . . . . .
6.3.3 Synthetic Aperture Radar (SAR) . . . . .
Distortions in radar images . . . . . . . . . . . .
6.4.1 Scale distortions . . . . . . . . . . . . . .
6.4.2 Terrain-induced distortions . . . . . . . .
6.4.3 Radiometric distortions . . . . . . . . . .
Interpretation of radar images . . . . . . . . . . .
6.5.1 Microwave signal and object interactions
6.5.2 Scattering patterns . . . . . . . . . . . . .
6.5.3 Applications of radar . . . . . . . . . . . .
Advanced radar processing techniques . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.

219
C. V. Reeves
. . . . . 220
. . . . . 221
. . . . . 223
. . . . . 227
. . . . . 229

Remote sensing below the ground surface

7.1
7.2
7.3
7.4
7.5

first

Introduction . . . . . . . . . . . . . . . . .
Gamma-ray surveys . . . . . . . . . . . .
Gravity and magnetic anomaly mapping
Electrical imaging . . . . . . . . . . . . . .
Seismic surveying . . . . . . . . . . . . . .

last

back

exit

.
.
.
.
.

zoom

.
.
.
.
.

contents

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

184
C. Pohl
. . 185
. . 187
. . 192
. . 193
. . 195
. . 198
. . 199
. . 200
. . 201
. . 205
. . 209
. . 210
. . 213
. . 214
. . 215

index

about

Contents
8

Radiometric aspects
8.1
8.2

8.3

Introduction . . . . . . . . . . . . . .
Cosmetic corrections . . . . . . . . .
8.2.1 Periodic line dropouts . . . .
8.2.2 Line striping . . . . . . . . . .
8.2.3 Random noise or spike noise
Atmospheric Corrections . . . . . . .
8.3.1 Haze correction . . . . . . . .
8.3.2 Sun angle correction . . . . .
8.3.3 Skylight correction . . . . . .

.
.
.
.
.
.
.
.
.

252
L. L. F. Janssen & M. J. C. Weir
. . . . . . . . . . . . . . . . 253
. . . . . . . . . . . . . . . . 255
. . . . . . . . . . . . . . . . 258
. . . . . . . . . . . . . . . . 259
. . . . . . . . . . . . . . . . 262
. . . . . . . . . . . . . . . . 265
. . . . . . . . . . . . . . . . 266
. . . . . . . . . . . . . . . . 268
. . . . . . . . . . . . . . . . 269

Geometric aspects
9.1
9.2
9.3

9.4

Introduction . . . . . . . . . . .
Relief displacement . . . . . . .
Two-dimensional approaches .
9.3.1 Georeferencing . . . . .
9.3.2 Geocoding . . . . . . . .
Three-dimensional approaches
9.4.1 Monoplotting . . . . . .
9.4.2 Orthoimage production
9.4.3 Stereoplotting . . . . . .

.
.
.
.
.
.
.
.
.

235
A. Prakash
. . . . 236
. . . . 237
. . . . 238
. . . . 240
. . . . 242
. . . . 243
. . . . 244
. . . . 245
. . . . 247

10 Image enhancement and visualisation

275
B. G. H. Gorte
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

first

last

back

exit

zoom

contents

index

about

Contents
10.2 Perception of colour . . . . . . . . . . . . . . . . . . . .
10.2.1 Tri-stimuli model . . . . . . . . . . . . . . . . .
10.2.2 Colour spaces . . . . . . . . . . . . . . . . . . .
10.3 Visualization of image data . . . . . . . . . . . . . . .
10.3.1 Histograms . . . . . . . . . . . . . . . . . . . .
10.3.2 Single band image display . . . . . . . . . . . .
10.4 Colour composites . . . . . . . . . . . . . . . . . . . .
10.4.1 Application of RGB and IHS for image fusion
10.5 Filter operations . . . . . . . . . . . . . . . . . . . . . .
10.5.1 Noise reduction . . . . . . . . . . . . . . . . . .
10.5.2 Edge enhancement . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

306
L. L. F. Janssen
. . . . . . 307
. . . . . . 309
. . . . . . 310
. . . . . . 312
. . . . . . 315
. . . . . . 317
. . . . . . 318
. . . . . . 325
. . . . . . 331
. . . . . . 333

11 Visual image interpretation

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . .
11.2 Image understanding and interpretation . . . . . . .
11.2.1 Human vision . . . . . . . . . . . . . . . . . .
11.2.2 Interpretation elements . . . . . . . . . . . .
11.2.3 Stereoscopic vision . . . . . . . . . . . . . . .
11.3 Application of visual image interpretation . . . . . .
11.3.1 Soil mapping with aerial photographs . . . .
11.3.2 Land cover mapping from multispectral data
11.3.3 Some general aspects . . . . . . . . . . . . . .
11.4 Quality aspects . . . . . . . . . . . . . . . . . . . . .
12 Digital image classification

first

last

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.

277
278
280
286
287
290
293
295
297
299
300

341
L. L. F. Janssen & B. G. H. Gorte

back

exit

zoom

contents

index

about

Contents
12.1 Introduction . . . . . . . . . . . . . . . . . .
12.2 Principle of image classification . . . . . . .
12.2.1 Image space . . . . . . . . . . . . . .
12.2.2 Feature space . . . . . . . . . . . . .
12.2.3 Image classification . . . . . . . . . .
12.3 Image classification process . . . . . . . . .
12.3.1 Preparation for image classification
12.3.2 Supervised image classification . . .
12.3.3 Unsupervised image classification .
12.3.4 Classification algorithms . . . . . . .
12.4 Validation of the result . . . . . . . . . . . .
12.5 Problems in image classification . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.

342
344
345
346
349
351
353
355
356
359
365
368

Glossary

379

A SI units & prefixes

409

first

last

back

exit

zoom

contents

index

about

List of Figures
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8

Principle of ground-based methods . . . . .

Principle of remote sensing based methods
Remote sensing and ground-based methods
Sea surface temperature map . . . . . . . .
Ocean biomass map . . . . . . . . . . . . . .
Sea surface height map . . . . . . . . . . . .
Ocean surface wind map . . . . . . . . . . .
Structure of the textbook . . . . . . . . . . .

.
.
.
.
.
.
.
.

28
28
34
35
37
38
39
43

2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8

A remote sensing sensor measures energy . . . . . . . . . . . .

Electric and magnetic vectors of an electromagnetic wave . . .
Relationship between wavelength, frequency and energy . . .
Blackbody radiation curves based on Stefan-Boltzmanns law
The electromagnetic spectrum . . . . . . . . . . . . . . . . . . .
Energy interactions in the atmosphere and on the land . . . . .
Atmospheric transmission expressed as percentage . . . . . .
Electromagnetic spectrum of the Sun . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.

50
53
54
56
58
62
63
64

first

last

back

exit

zoom

.
.
.
.
.
.
.
.

contents

.
.
.
.
.
.
.
.

index

about

List of Figures
2.9
2.10
2.11
2.12
2.13
2.14
2.15

Rayleigh scattering . . . . . . . . . . . . . . . .
Rayleigh scattering affects the colour of the sky
Effects of clouds in optical remote sensing . . .
Specular and diffuse reflection . . . . . . . . . .
Reflectance curve of vegetation . . . . . . . . .
Reflectance curves of soil . . . . . . . . . . . . .
Reflectance curves of water . . . . . . . . . . .

.
.
.
.
.
.
.

66
67
69
70
73
75
76

3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11

Overview of sensors . . . . . . . . . . . . . . .
Example video image . . . . . . . . . . . . . . .
Example multispectral image . . . . . . . . . .
TSM derived from imaging spectrometer data
Example thermal image . . . . . . . . . . . . .
Example microwave radiometer image . . . . .
DTM derived by laser scanning . . . . . . . . .
Example radar image . . . . . . . . . . . . . . .
Roll, pitch and yaw angles . . . . . . . . . . . .
Meteorological observation system . . . . . . .
An image file comprises a number of bands . .

.
.
.
.
.
.
.
.
.
.
.

86
90
91
93
95
96
98
100
102
106
107

4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8

Vertical and oblique photography . . . . . . . . . . . . . . .

Vertical and oblique aerial photo of ITC building . . . . . .
Major components of an aerial camera . . . . . . . . . . . .
Lens cone of an aerial camera . . . . . . . . . . . . . . . . .
Auxiliary data annotation on an aerial photograph . . . . .
Section of a photographic film showing the emulsion layer
Spectral sensitivity curves . . . . . . . . . . . . . . . . . . .
Negative film and positive print . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.

120
121
122
124
125
128
131
133

first

last

back

exit

zoom

contents

index

about

List of Figures
4.9
4.10
4.11
4.12
4.13

Emulsion layers of a true colour film . . . . . .

True world, negative and positive . . . . . . . .
Comparison of different types of photography
Effect of focal length . . . . . . . . . . . . . . .
A survey area for block photography . . . . . .

.
.
.
.
.

134
135
137
140
144

5.1
5.2
5.3
5.4

Principle of the whiskbroom scanner . . . . . . . . . .

Spectral response curve of a sensor . . . . . . . . . . .
Principle of a pushbroom scanner . . . . . . . . . . . .
NOAA/AVHRR spatial resolution varies significantly

.
.
.
.

156
158
160
167

6.1
6.2
6.3
6.4
6.5
6.6
6.7

Principle of active microwave remote sensing . . . . . . . . . . .

Microwave spectrum and band identification by letters . . . . .
Polarization of electromagnetic waves . . . . . . . . . . . . . . .
Radar remote sensing geometry . . . . . . . . . . . . . . . . . . .
Radar incidence angle and local incidence angle . . . . . . . . .
Geometric distortions in radar imagery due to terrain elevations
Original and speckle filtered radar image . . . . . . . . . . . . .

186
190
191
192
193
194
205

7.1
7.2
7.3
7.4
7.5

Abundance map of K, Th and U from gamma-ray measurements

Sea floor topography as determined by satellite altimetry . . . .
Magnetic anomaly map derived by an airborne magnetometer .
Conductivity measured by airborne measurements . . . . . . . .
3D terrain from seismic surveys . . . . . . . . . . . . . . . . . . .

222
224
226
228
230

8.1
8.2
8.3

Image pre-processing steps . . . . . . . . . . . . . . . . . . . . . 236

Simulated Landsat MSS image . . . . . . . . . . . . . . . . . . . 237
Image with line-dropouts . . . . . . . . . . . . . . . . . . . . . . 238

first

last

back

exit

zoom

.
.
.
.
.

contents

index

about

List of Figures
8.4
8.5
8.6
8.7

Image corrected for line-dropouts

Image with line striping . . . . .
Image with spike noise . . . . . .
Variations of the sun angle . . . .

.
.
.
.

239
240
242
245

9.1
9.2
9.3
9.4
9.5
9.6
9.7

Illustration of the effect of terrain topography . . . . .

Illustration of height displacement . . . . . . . . . . .
Image and map coordinate systems . . . . . . . . . . .
Resampling methods . . . . . . . . . . . . . . . . . . .
Original, georeferenced and geocoded satellite image
The process of digital monoplotting . . . . . . . . . .
Illustration of parallax in stereo pair . . . . . . . . . .

.
.
.
.
.
.
.

255
257
258
263
264
266
270

10.1
10.2
10.3
10.4
10.5
10.6
10.7
10.8
10.9
10.10
10.11
10.12

Sensitivity curves of the human eye . . . . . . . . . . . . . . . . .

Comparison of additive and subtractive colour schemes . . . . .
The RGB cube . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Relationship between RGB and IHS colour spaces . . . . . . . .
Standard and cumulative histogram . . . . . . . . . . . . . . . .
Multi-band image displayed on a monitor . . . . . . . . . . . . .
Transfer functions . . . . . . . . . . . . . . . . . . . . . . . . . . .
Single band display . . . . . . . . . . . . . . . . . . . . . . . . . .
Multi-band display . . . . . . . . . . . . . . . . . . . . . . . . . .
Procedure to merge SPOT panchromatic and multispectral data
Input and output of a filter operation . . . . . . . . . . . . . . . .
Original, edge enhanced and smoothed image . . . . . . . . . .

278
281
282
283
289
290
291
292
294
296
297
300

11.1
11.2

Satellite image of Antequera area in Spain . . . . . . . . . . . . . 310

Mud huts of Labbezanga near the Niger river . . . . . . . . . . . 311

first

last

back

exit

.
.
.
.

zoom

.
.
.
.

contents

index

about

List of Figures
11.3
11.4
11.5
11.6
11.7
11.8
11.9

The mirror stereoscope . . . . . . . . . . . .

Panchromatic photograph to be interpreted
Photo-interpretation transparency . . . . .
Land use and land cover. . . . . . . . . . . .
Sample of the CORINE land cover map. . .
Comparison of different line maps . . . . .
Comparison of different thematic maps . .

.
.
.
.
.
.
.

316
319
320
326
327
334
335

12.1
12.2
12.3
12.4
12.5
12.6
12.7
12.8
12.9
12.10
12.11
12.12

The structure of a multi-band image . . . . . . . .

Two- and three-dimensional feature space . . . . .
Scatterplot of a digital image . . . . . . . . . . . . .
Distances in the feature space . . . . . . . . . . . .
Feature space showing six clusters of observations
The classification process . . . . . . . . . . . . . . .
Image classification input and output . . . . . . .
Results of a clustering algorithm . . . . . . . . . .
Box classification . . . . . . . . . . . . . . . . . . .
Minimum distance to mean classification . . . . .
Maximum likelihood classification . . . . . . . . .
The mixed pixel or mixel . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.

345
346
347
348
350
351
352
358
360
362
364
370

first

last

back

exit

zoom

.
.
.
.
.
.
.

contents

index

about

List of Tables
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9

Meteosat-5 VISSR characteristics . . . . . .

NOAA-15 AVHRR characteristics . . . . . .
Landsat-7 ETM+ characteristics . . . . . . .
Principal applications of Landsat TM bands
SPOT-4 HRVIR characteristics . . . . . . . .
IRS-1D PAN characteristics . . . . . . . . .
IKONOS OSA characteristics . . . . . . . .
Terra characteristics . . . . . . . . . . . . . .
EO-1 characteristics . . . . . . . . . . . . . .

9.1

Sample set of ground control points . . . . . . . . . . . . . . . . . 260

10.1
10.2
10.3
10.4
10.5

Example histogram in tabular format . .

Summary histogram statistics . . . . . .
Filter kernel for smoothing . . . . . . . .
Filter kernel for weighted smoothing . .
Filter kernel used for edge enhancement

first

last

back

exit

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

zoom

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

contents

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

index

.
.
.
.
.
.
.
.
.

.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.

165
166
168
170
171
172
173
176
178

288
289
299
299
300

about

List of Tables
11.1
11.2
11.3
11.4
11.5

Example geopedologic legend . . . . . . . . . . . .

Example soil legend . . . . . . . . . . . . . . . . . .
CORINE land cover classes . . . . . . . . . . . . .
CORINE land cover classes (2) . . . . . . . . . . .
CORINEs description of Mineral Extraction Sites

.
.
.
.
.

323
324
328
329
330

12.1 Sample error matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

12.2 Spectral, land cover and land use classes . . . . . . . . . . . . . . . 369
A.1
A.2
A.3
A.4

first

Relevant SI units in the context of remote sensing

Unit prefix notation . . . . . . . . . . . . . . . . . .
Common units of wavelength . . . . . . . . . . . .
Constants and non-SI units . . . . . . . . . . . . .

last

back

exit

zoom

.
.
.
.

contents

.
.
.
.

index

.
.
.
.

409
410
410
410

about

Preface
Principles of Remote Sensing is the basic textbook on remote sensing for all students enrolled in the 20002001 educational programmes at ITC. As well as
being a basic textbook for the institutes regular MSc and PM courses, Principles
of Remote Sensing will be used in various short courses and possibly also by ITCs
sister institutes. The first edition is an extensively revised version of an earlier
text produced for the 19992000 programme. Principles of Remote Sensing and the
companion volume, Principles of Geographic Information Systems [6], are published
in the ITC Educational Textbook series. We need to go back to the 1960s to find
a similar official ITC textbook on subjects related to Remote Sensing: the ITC
Textbooks on Photogrammetry and Photo-interpretation, published in English
and French [11, 12].
You may wonder why ITC has now produced its own introductory textbook while there are already many books on the subject available on the market. Principles of Remote Sensing is different in various aspects. First of all, it
has been developed for the specific ITC student population, thereby taking into
account their entry level and knowledge of English language. The textbook relates to the typical ITC application disciplines and among others provides an
introduction into techniques which acquire sub-surface characteristics. As the
first

last

back

exit

zoom

contents

index

about

Preface
textbook is used in the start of the programmes, it tries to stimulate conceptual
and abstract thinking by providing and explaining some fundamental, however
simple, equations (in general, no more than one equation per chapter). Principles of Remote Sensing aims to provide a balanced approach towards traditional
photogrammetric and remote sensing subjects: three sensors (aerial camera, multispectral scanner and radar) are dealt with more or less the same detail. Finally,
compared to other introductory textbooks which often focus on the technique,
Principles of Remote Sensing also introduces processes. In this sense, it provides
a frame to refer to when more detailed subjects are dealt with later in the programme.

first

last

back

exit

zoom

contents

index

about

Preface

How to use the material

Principles of Remote Sensing has been produced both as a hard-copy textbook and
as an electronic document. In this way, the student is offered the optimal combination to study the subject, and, to use the book as a general reference. Each
chapter gives a summary and provides questions for self test. The book comprises a glossary, an index and a bibliography. The electronic document (PDF
format) enables fast navigation and quick referencing.

first

last

back

exit

zoom

contents

index

about

Preface

Acknowledgements
This textbook is the result of a process to define and develop material for a core
curriculum. This process started in 1998 and was carried out by a working group
comprising of Rolf de By, Michael Weir, Cees van Westen, myself, chaired by
Ineke ten Dam and supported by Erica Weijer. This group put many efforts
in the definition and realization of the earlier version of the two core textbooks.
Ineke was also supervising the process leading to this result. My fellow working
group members are greatly acknowledged for their support.
This textbook could not have materialized without the efforts of the (co) authors of the chapters: Wim Bakker, Ben Gorte, John Horn, Christine Pohl, Colin
Reeves, Michael Weir and Tsehaie Woldai. Many other colleagues contributed
one way or another to either the earlier version or this version of Principles of Remote Sensing: Paul Hofstee, Gerrit Huurneman, Yousif Hussin, David Rossiter,
Rob Soeters, Ernst Schetselaar, Andrew Skidmore, Dhruba Shrestha and Zoltan
Vekerdy.
The design and implementation of the textbook layout, of both the hard-copy
and electronic document, is the work of Rolf de By. Using the LATEX typesetting system, Rolf realized a well structured and visually attractive document to
study. Many of the illustrations in the book have been provided by the authors,
supported by Job Duim and Gerard Reinink. Final editing of the illustrations
was done by Wim Feringa who also designed the cover.
Michael Weir has done a tremendous job in checking the complete textbook
on English spelling and grammar. We know that our students will profit from
this.
The work on this textbook was greatly stimulated through close collaboration

first

last

back

exit

zoom

contents

index

about

Preface
with the editor of Principles of Geographic Information Systems, Rolf de By.
Lucas L. F. Janssen, Enschede, September 2000

first

last

back

exit

zoom

contents

index

about

Preface

Preface to the second edition

After Lucas Janssen left ITC last year, moving into another job, I was asked to
be the editor for the next version(s) of this book. I got impressed about the work
Lucas, in cooperation with many others, achieved.
The LATEX frame, developed by Rolf de By is of a clear construction. It enabled
me to enter new items and to modify text, equations and other objects without
being familiar with this typesetting system. However, without the permanent
support of Rolf I could never have finished this version.
Following many constructive suggestions received from users and reviewers,
for which I am very grateful, a number of changes were realized. Especially
I thank Wim Bakker for his careful reading of all chapters and his suggestions
to improve the book and Wim Feringa for his contribution to the creation and
modification of a number of illustrations.
The major changes made in this version are:
A chapter on Radiometric aspects is included, treating the main distortions
caused by sensors and/or the atmosphere. The text and the corresponding figures were produced by my colleague Anupma Prakash and she is
acknowledged for her contribution.
A number of new sensor systems is added to Chapter 5 on Multispectral
scanners.
On demand of the ITC working group Core Modules, the Chapters 6
and 7 are excluded from the ITC core modules, and also a number of
(sub)sections. They are indicated by a special icon in the margin, depicting
a dangerous bend road sign.

first

last

back

exit

zoom

contents

index

about

Preface
Gerrit C. Huurneman, Enschede, August 2001

first

last

back

exit

zoom

contents

index

about

Chapter 1
Introduction to remote sensing

first

last

back

exit

zoom

contents

index

about

1.1. Spatial data acquisition

1.1

Spatial data acquisition

All ITC students, one way or another, deal with georeferenced data. They might
be involved in the collection of data, processing of the data, analysis of the data
or actually using the data for decision making. In the end, data are acquired
to yield information for management purposes: water management, land management, resources management, et cetera. By data we mean representations that
can be operated upon by a computer; by information we mean data that has been
interpreted by human beings (Principles of GIS, Chapter 1). This textbook focusses on the methods used to collect georeferenced, or geo-spatial, data. The
need for spatial data is best illustrated by some examples.
An agronomist is interested in forecasting the overall agricultural production of a large area. This requires data on the area planted with different
crops and data on biomass production to estimate the yield.
An urban planner needs to identify areas in which dwellings have been
built illegally. The different types of houses and their configuration needs
to be determined. The information should be in a format that enables integration with other socio-economic information.
An engineer needs to determine the optimal configuration for siting of relay stations for a telecommunication company. The optimal configuration
primarily depends on the form of the terrain and on the location of obstacles such as buildings.
A mining engineer is asked to explore an area and to provide a map of the
surface mineralogy. In addition, s/he should start to give a first estimation
of the effect of water pumping on the neighbouring agricultural region.
first

last

back

exit

zoom

contents

index

about

1.1. Spatial data acquisition

phe A climatologist would like to understand the causes of the El Nino
nomenon. For this, s/he would need data on many parameters including
sea currents, sea surface temperature, sea level, meteorological parameters,
energy interactions between the land and water surface, et cetera.
Note that all of the above examples deal with spatial phenomena; in fact,
with spatio-temporal phenomena since time is an important dimension too. To satisfy the information requirements of the above mentioned examples a wide variety of methods will be used: conducting interviews, land surveying, laboratory
measurements of samples, interpretation of satellite images, measurements by in
situ sensors, using aerial photographs, running numerical models, et cetera. For
our purposes, it makes sense to distinguish between ground-based and remote
sensing methods.

first

last

back

exit

zoom

contents

index

about

1.1. Spatial data acquisition

Ground-based and Remote Sensing Methods
In principle, there are two main categories of spatial data acquisition:
ground-based methods such as making field observations, taking in situ measurements and performing land surveying. Using ground-based methods,
you operate in the real world environment (Figure 1.1).
Observation and
measurement

Real world

Spatial
database

Figure 1.1: The principle of a ground-based

method: measurements
and observations are performed in the real world.

Ground based approach

remote sensing methods, which are based on the use of image data acquired
by a sensor such as aerial cameras, scanners or a radar. Taking a remote
sensing approach means that information is derived from the image data,
which form a (limited) representation of the real world (Figure 1.2).
Real world

Sensor

Image data

Observation and
measurement

Spatial
database

Figure 1.2: The principle of a remote sensing

1.1. Spatial data acquisition

Another term to introduce here is Earth Observation (EO), which usually refers
to spaceborne remote sensing.
Before the image data can yield the required information about the objects or
phenomena of interest, they need to be processed. The analysis and information
extraction or information production is part of the overall remote sensing process.
In this textbook, remote sensing refers to all aerospace remote sensing techniques, including aerial photography. Sometimes, because of historical reasons,
aerial photos are referred to as a separate aspect of aerospace survey.

first

Remote sensing requires ground data.
Although remote sensing data can be interpreted and processed without other
information, the best results are obtained by linking remote sensing measurements to ground (or surface) measurements and observations. Highly accurate
SST maps can be derived from satellite image data when combined with in situ
temperature measurements (by buoy). This idea of complementarity between
remote sensing and ground-based surveying is also part of ITCs name: aerospace
surveying. Figure 1.3 illustrates this concept.

first

Real world

RS sensor

last

Image data

back

exit

Analysis

zoom

Figure 1.3: In most situations, remote sensing

based data acquisition is
complemented by groundbased measurements and
observations.

Spatial
database

contents

index

about

1.2. Application of remote sensing

Remote sensing provides area covering data
The spatial pattern of SST can best be studied using map-like products. These
can be obtained by interpolation from in situ measurements by buoys. However,
it is unlikely that the number and distribution of buoys is adequate to enable
detection/identification of relevant surface features. The distance between the
is in the order of 2000 km (in x-direction) and 300 km
buoys network for El Nino
(in y-direction). Using data of meteorological satellites, SST can be assessed at a
level of detail up to 1 km (Figure 1.4). RS derived SST data can, for example, be
used to determine an optimal density and location of buoys.

Figure 1.4:
Sea surface
temperature
as
determined from NOAAAVHRR data. Courtesy of
NOAA

first

last

back

exit

zoom

contents

index

about

Remote sensing provides multipurpose image data

In this example, SST maps are used to study El Nino.

The same data could
be used years later to find, for example, a relation between SST and the algae
blooms around Pacific islands. SST maps are not only of interest to researchers
but also to large fishing companies that want to guide their vessels to promising
fishing grounds. A data set, thus, can be of use for more than one organization
and can also prove useful for historical studies.

first

last

back

exit

zoom

contents

index

about

1.2. Application of remote sensing

Remote sensing is cost-effective
The validity of this statement is sometimes hard to assess, especially when dealing with spaceborne remote sensing. Consider an international scientific project
phenomenon. Installation and maintenance of buoys
that studies the El Nino
costs a lot of money. Meteorological satellites have already been paid for and the
data can be considered free. Remote sensing would thus be a cheap technique.
However, as RS can only be used to measure surface characteristics, buoys still
need to be placed to determine subsurface characteristics.
Although the above statements are related to a scientific problem of a global
phenomenon, they are generic to other application contexts. Consider the above
statements related to the monitoring of (illegal) urban growth around African
cities using aerial photography.
Probably, you will conclude that you do not have all the knowledge required
to comment on these statements in the urban context. You may even find it
difficult to identify which remote sensing technique solves your own particular information requirements. After studying this textbook Principles of Remote
Sensing you may expect to be better equipped to consider these issues.

first

last

back

exit

zoom

contents

index

about

1.3. Structure of this textbook

Real world

Sensor

Image data

Observation and
measurement

Spatial
database

2- Electromagnetic
energy
34567-

Sensors and platforms

Aerial camera
Multispectral scanner
Radar
RS below the ground
surface
8- Radiometric aspects
9- Geometric aspects

Figure 1.8: Relationship

between the chapters of
this textbook and the remote sensing process

10- Image enhancement and visualisation

contents

index

about

1.3. Structure of this textbook

example, outline how the statements about RS relate

4. Following the El Nino
to a particular case of your organization (e.g., land resources, geology, et
cetera).

first

last

back

exit

zoom

contents

index

about

1.3. Structure of this textbook

These are typical exam questions:
1. Explain, or give an example, how ground-based and remote sensing methods may complement each other.
2. List three possible limitations of remote sensing data.

first

last

back

exit

zoom

contents

index

about

Chapter 2
Electromagnetic energy and remote
sensing

first

last

back

exit

zoom

contents

index

about

2.1. Introduction

2.1

Introduction

Remote sensing relies on the measurement of electromagnetic (EM) energy. EM

energy can take several different forms. The most important source of EM energy at the Earths surface is the Sun, which provides us, for example, with (visible) light, heat (that we can feel) and UV-light, which can be harmful to our
skin.
Sun

Passive
Sensor

Reflected
Sunlight

Active
Sensor

Earth
energy

Figure 2.1:
A remote
sensing sensor measures
reflected or emitted energy. An active sensor has
its own source of energy.

Earths surface

Many sensors used in remote sensing measure reflected sunlight. Some sensors, however, detect energy emitted by the Earth itself or provide their own
energy (Figure 2.1). A basic understanding of EM energy, its characteristics and
its interactions is required to understand the principle of the remote sensor. This
knowledge is also needed to interpret remote sensing data correctly. For these
reasons, this chapter introduces the basic physics of remote sensing.
In Section 2.2, EM energy, its source and the different parts of the electrofirst

last

back

exit

zoom

contents

index

about

2.1. Introduction
magnetic spectrum are explained. In between the remote sensor and the Earths
surface is the atmosphere that influences the energy that travels from the Earths
surface to the sensor. The main interactions between EM waves and the atmosphere are described in Section 2.3. Section 2.4 introduces the interactions that
take place at the Earths surface.

first

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

2.2

first

Electromagnetic energy

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

2.2.1 Waves and photons

Electromagnetic (EM) energy can be modelled in two ways: by waves or by
energy bearing particles called photons. In the wave model, electromagnetic energy is considered to propagate through space in the form of sine waves. These
waves are characterized by two fields, electrical (E) and magnetic (M) fields,
which are perpendicular to each other. For this reason, the term electromagnetic
energy is used. The vibration of both fields is perpendicular to the direction of
travel of the wave (Figure 2.2). Both fields propagate through space at the speed
of light c, which is 299,790,000 m/s and can be rounded off to 3 108 m/s.
Electric field

Wavelength, l

Distance
Magnetic field

Figure 2.2: Electric (E)

and magnetic (M) vectors of an electromagnetic
wave.

Velocity of light, c

One characteristic of electromagnetic waves is particularly important for understanding remote sensing. This is the wavelength that is defined as the distance between successive wave crests (Figure 2.2). Wavelength is measured in
metres (m) or some factor of metres such as nanometres (nm, 109 m) or micrometres (m, 106 m). (For an explanation of units and prefixes refer to Appendix 1).
The frequency, v, is the number of cycles of a wave passing a fixed point
over a specific period of time. Frequency is normally measured in hertz (Hz),
first

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

which is equivalent to one cycle per second. Since the speed of light is constant,
wavelength and frequency are inversely related to each other:
(2.1)

c = v.

In this equation, c is the speed of light (3 108 m/s), is the wavelength (m), and
v is the frequency (cycles per second, Hz).
The shorter the wavelength, the higher the frequency. Conversely, the longer
the wavelength, the lower the frequency (Figure 2.3).
Short wavelength

Long wavelength

High frequency
High energy

Figure 2.3: Relationship

between wavelength, frequency and energy

Low frequency
Low energy

Most characteristics of EM energy can be described using the wave model as

described above. For some purposes, however, EM energy is more conveniently
modelled by the particle theory, in which EM energy is composed of discrete
units called photons. This approach is taken when quantifying the amount of
energy measured by multispectral sensor (Section 5.2.1). The amount of energy
held by a photon of a specific wavelength is then given by
c
Q=hv =h ,
(2.2)

where Q is the energy of a photon (J), h is Plancks constant (6.6262 1034 J s),
and v the frequency (Hz). From Equation 2.2 it follows that the longer the wavelength, the lower its energy content. Gamma rays (around 109 m) are the most
first

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

energetic, and radio waves (> 1 m) the least energetic. An important consequence for remote sensing is that it is more difficult to measure the energy emitted in longer wavelengths than in shorter wavelengths.

first

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

2.2.2 Sources of EM energy

If T increases, peak moves

towards shorter wavelengths,
and the total area under the
curve increases

4
12
T=

Spectral radiant emittance (W/cm2)/m

All matter with a temperature above absolute zero (0 K, where n C = n + 273 K)

radiates EM energy due to molecular agitation. Agitation is the movement of
the molecules. This means that the Sun, and also the Earth, radiate energy in
the form of waves. Matter that is capable of absorbing and re-emitting all EM
energy is known as a blackbody. For blackbodies both the emissivity, , and the
absorptance, , are equal to (the maximum value of) 1.
The amount of energy radiated by an object depends on its absolute temperature, its emissivity and is a function of the wavelength. In physics, this principle
is defined by Stefan-Boltzmanns Law. A blackbody radiates a continuum of

107

T=873

Figure 2.4:
Blackbody
radiation curves based on
Stefan-Boltzmanns
law
(with temperatures in K)

T=673

1.0

first

2.0

3.0

last

4.0

back

5.0

exit

6.0

zoom

7.0

contents

8.0

index

about

2.2. Electromagnetic energy

wavelengths. The radiation emitted by a blackbody at different temperatures
is shown in Figure 2.4. Note the units in this figure: the x-axis indicates the
wavelength and the y-axis indicates the amount of energy per unit area. The
area below the curve, therefore, represents the total amount of energy emitted
at a specific temperature. From Figure 2.4 it can be concluded that a higher
temperature corresponds to a greater contribution of shorter wavelengths. The
peak radiation at 400 C is around 4 m while the peak radiation at 1000 C is at
2.5 m. The emitting ability of a real material compared to that of the blackbody
is referred to as the materials emissivity. In reality, blackbodies are hardly found
in nature; most natural objects have emissivities less than one. This means that
only part, usually between 8098%, of the received energy is re-emitted. Consequently, part of the energy is absorbed. This physical property is relevant in, for
example, the modelling of global warming processes.

first

last

back

exit

zoom

contents

index

about

2.2. Electromagnetic energy

2.2.3 Electromagnetic spectrum

All matter with a certain temperature radiates electromagnetic waves of various
wavelengths. The total range of wavelengths is commonly referred to as the
electromagnetic spectrum (Figure 2.5). It extends from gamma rays to radio
waves.
0.6

0.7 (mm)

red

green

0.5

blue

0.4

Near-infrared

Visible

Wavelength (mm)

10-6 10-5 10-4 10-3 10-2 10-1

102

103

104 105

106

107 108

109

d
an io
d
ra

l
Te

al
m
er
Th -IR
id -IR

M ar
V
e le (U
N ib let
s o
Viravi

Figure 2.5: The electromagnetic spectrum, after

[16]

2.3. Energy interaction in the atmosphere

2.3

Energy interaction in the atmosphere

The most important source of energy is the Sun. Before the Suns energy reaches
the Earths surface, three fundamental interactions in the atmosphere are possible: absorption, transmission and scattering. The energy transmitted is then
reflected or absorbed by the surface material. (Figure 2.6).

first

last

back

exit

zoom

contents

index

about

2.3. Energy interaction in the atmosphere

Sun

y
nerg
ent e

Incid
RS
Sensor

Atmospheric
absorbtion
Scattered
radiation

Cloud

Direct
radiation

Scattered
radiation

Earth

first

Atmospheric
emission
Thermal
emission

Reflected
radiation

Reflection
processes

last

back

Figure 2.6: Energy interactions in the atmosphere

and on the land

Emission
processes

exit

zoom

contents

index

about

2.3. Energy interaction in the atmosphere

2.3.1 Absorption and transmission

Electromagnetic energy travelling through the atmosphere is partly absorbed by
various molecules. The most efficient absorbers of solar radiation in the atmosphere are ozone (O3 ), water vapour (H2 O) and carbon dioxide (CO2 ).
% transmission

100

H2O
and
Co2

Co2

H2O

Co2

H2O

Figure 2.7: Atmospheric

transmission expressed as
percentage

Wavelength, microns

Figure 2.7 gives a schematic representation of the atmospheric transmission

in the 022 m wavelength region. From this figure it may be seen that about
half of the spectrum between 022 m is useless for remote sensing of the Earths
surface, simply because none of the corresponding energy can penetrate the atmosphere. Only the wavelength regions outside the main absorption bands of
the atmospheric gases can be used for remote sensing. These regions are referred
to as the atmospheric transmission windows and include:
A window in the visible and reflected infrared region, between 0.42 m.
This is the window where the (optical) remote sensors operate.
Three windows in the thermal infrared region, namely two narrow windows around 3 and 5 m, and a third, relatively broad, window extending
from approximately 8 to 14 m.
first

last

back

exit

zoom

contents

index

about

2.3. Energy interaction in the atmosphere

1.0 x
104
Solar extra terrestrial
6000 K blackbody
Solar at Earths surface

0.8
O3
O2
H2O
O2
H2 O

0.6
0.4

Visible

Spectral radiant emittance (W/cm2)/m

Because of the presence of atmospheric moisture, strong absorption bands are

found at longer wavelengths. There is hardly any transmission of energy in the
region from 22 m to 1 mm. The more or less transparent region beyond 1 mm
is the microwave region.

0.2
0.2

0.4

H2O

0.8

1.0

1.8 2.0

1.4

Figure 2.8: The electromagnetic spectrum of the

Sun as observed with and
without the influence of the
Earths atmosphere.

H2O
CO2

H2O

2.4

2.6

2.8

3.0

Wavelength m

The solar spectrum as observed both with and without the influence of the
Earths atmosphere is shown in Figure 2.8. First of all, look at the radiation curve
of the Sun (measured outside the influence of the Earths atmosphere), which
resembles a blackbody curve at 6000 K. Secondly, compare this curve with the
radiation curve as measured at the Earths surface. The relative dips in this curve
indicate the absorption by different gases in the atmosphere.

first

last

back

exit

zoom

contents

index

about

2.3. Energy interaction in the atmosphere

about

2.3. Energy interaction in the atmosphere

Mie scattering
Mie scattering occurs when the wavelength of the incoming radiation is similar
in size to the atmospheric particles. The most important cause of Mie scattering
are the aerosols: a mixture of gases, water vapour and dust.
Mie scattering is generally restricted to the lower atmosphere where larger
particles are more abundant, and dominates under overcast cloud conditions.
Mie scattering influences the entire spectral region from the near-ultraviolet up
to and including the near-infrared.

first

last

back

exit

zoom

contents

index

about

2.3. Energy interaction in the atmosphere

Non-selective scattering
Non-selective scattering occurs when the particle size is much larger than the radiation wavelength. Typical particles responsible for this effect are water droplets
and larger dust particles.
Non-selective scattering is independent of wavelength, with all wavelengths
scattered about equally. The most prominent example of non-selective scattering includes the effect of clouds (clouds consist of water droplets). Since all
wavelengths are scattered equally, a cloud appears white. Optical remote sensing cannot penetrate clouds. Clouds also have a secondary effect: shadowed
regions on the Earths surface (Figure 2.11).
Sensor
Su

nr
ad

Cloud

ow zo

Cloud

Figure 2.11: Direct and indirect effects of clouds in

optical remote sensing

Shadow

Earth

first

Zone of no
penetration

Shad

ce
an n
i
d
ra utio
d
b
ou tri
Cl con

ian

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

2.4

Energy interactions with the Earths surface

In land and water applications of remote sensing we are most interested in the
reflected radiation because this tells us something about surface characteristics.
Reflection occurs when radiation bounces off the target and is then redirected.
Absorption occurs when radiation is absorbed by the target. Transmission occurs when radiation passes through a target. Two types of reflection, which
represent the two extremes of the way in which energy is reflected from a target,
are specular reflection and diffuse reflection (Figure 2.12). In the real world, usually
a combination of both types is found.
Specular reflection, or mirror-like reflection, typically occurs when a surface is smooth and all (or almost all) of the energy is directed away from
the surface in a single direction. It is most likely to occur when the Sun is
high in the sky. Specular reflection can be caused, for example, by a water

(a)

first

last

Red

0
0.4

0.6

Colour IR
sensitive region

Green

Chlorophyll absorption

Blue
40

Leaf refle
cta

Percent reflectance

Chlorophyll absorption
Chlorophyll absorption

2.4. Energy interactions with the Earths surface

0.8

1.0

1.2
1.4
1.6
1.8
Wavelength, micrometers

2.0

2.2

2.4

Figure 2.13: An idealized

spectral reflectance curve
of a healthy vegetation

2.6

Vegetation
The reflectance characteristics of vegetation depend on the properties of the
leaves including the orientation and the structure of the leaf canopy. The proportion of the radiation reflected in the different parts of the spectrum depends
on leaf pigmentation, leaf thickness and composition (cell structure) and on the
amount of water in the leaf tissue. Figure 2.13 shows an ideal reflectance curve
from healthy vegetation. In the visible portion of the spectrum, the reflection
from the blue and red light is comparatively low since these portions are absorbed by the plant (mainly by chlorophyll) for photosynthesis and the vegetation reflects relatively more green light. The reflectance in the near-infrared is
highest but the amount depends on the leaf development and the cell structure
first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

of the leaves. In the middle infrared, the reflectance is mainly determined by
the free water in the leaf tissue; more free water results in less reflectance. They
are therefore called water absorption bands. When the leaves dry out, for an
example during the harvest time of the crops, the plant may change colour (for
example, to yellow). At this stage there is no photosynthesis, causing reflectance
in the red portion of the spectrum to be higher. Also, the leaves will dry out
resulting in higher reflectance in the middle infrared whereas the reflectance in
the near-infrared may decrease. As a result, optical remote sensing data provide
information about the type of plant and also about its health condition.

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

Bare soil
Surface reflectance from bare soil is dependent on so many factors that it is difficult to give one typical soil reflectance curve. However, the main factors influencing the soil reflectance are soil colour, moisture content, the presence of
carbonates, and iron oxide content. Figure 2.14 gives some reflectance curves
for the five main types of soil occurring in the USA. Note the typical shapes
of most of the curves, which show a convex shape between 5001300 nm with
dips at 1450 and 1950 nm. These dips are so-called water absorbtion bands and are
caused by the presence of soil moisture. The iron-dominated soil has a quite different reflectance curve that can be explained by the iron absorbtion dominating
at longer wavelengths.
40

reflectance (%)

b
c

Figure 2.14: Reflectance

spectra of surface samples of five mineral soils,
(a) organic dominated,
(b) minimally altered,
(c) iron altered, (d) organic
affected and (e) iron
dominated (from [17])

d
a

e
0
400

800

1200

1600

2000

2400

wavelength (nm)

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

Water
Compared to vegetation and soils, water has the low reflectance. Vegetation
may reflect up to 50%, soils, up to 3040% while water reflects at the most 10%
of the incoming radiation. Water reflects EM energy in the visible up to the nearinfrared. Beyond 1200 nm all energy is absorbed. Some curves of different types
of water are given in Figure 2.15 The highest reflectance is given by turbid (silt
loaded) water and water containing plants with a chlorophyll reflection peak at
the green wavelength.

reflectance (%)

30
20

Figure 2.15: Typical effects of chlorophyll and

sediments on water reflectance: (a) ocean water,
(b) turbid water, (c) water
with chlorophyll (from [17])

10
a

400

800

1200

1600

2000

2400

wavelength (nm)

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

Summary
Remote sensing is based on the measurement of Electromagnetic (EM) energy.
EM energy propagates through space in the form of sine waves characterized by
electrical (E) and magnetic (M) fields, which are perpendicular to each other. EM
can be modelled either by waves or by energy bearing particles called photons.
One property of EM waves that is particularly important for understanding remote sensing is the wavelength (), defined as the distance between successive
wave crests measured in metres (m), nanometres (nm, 109 m) or micrometres
(m, 106 m). The frequency is the number of cycles of a wave passing a fixed
point in a specific period of time and is measured in hertz (Hz). Since the speed
of light is constant, wavelength and frequency are inversely related to each other.
The shorter the wavelength, the higher the frequency and vice versa.
All matter with a temperature above the absolute zero (0 K) radiates EM
energy due to molecular agitation. Matter that is capable of absorbing and reemitting all EM energy received is known as a blackbody. All matter with a
certain temperature radiates electromagnetic waves of various wavelengths depending on its temperature. The total range of wavelengths is commonly referred to as the electromagnetic spectrum. It extends from gamma rays to radio
waves. The amount of energy detected by a remote sensing system is a function
of the interactions on the way to the object, the object itself and the interactions
on the way returning to the sensor.
The interactions of the Suns energy with physical materials, both in the atmosphere and at the Earths surface, cause this energy to be reflected, absorbed,
transmitted or scattered. Electromagnetic energy travelling through the atmosphere is partly absorbed by molecules. The most efficient absorbers of solar
radiation in the atmosphere are ozone (O3 ), water vapour (H2 O) and carbon

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

dioxide (CO2 ).
Atmospheric scattering occurs when the particles or gaseous molecules present in the atmosphere interact with the electromagnetic radiation and cause
it to be redirected from its original path. Three types of scattering take place:
Rayleigh scattering, Mie Scattering and Non-selective scattering.
When electromagnetic energy from the Sun hits the Earths surface, three
fundamental energy interactions are possible: absorption, transmission, and reflectance. Specular reflection occurs when a surface is smooth and all of the
energy is directed away from the surface in a single direction. Diffuse reflection
occurs when the surface is rough and the energy is reflected almost uniformly in
all directions.

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

Questions
The following questions can help you to study Chapter 2.
1. What are advantages/disadvantages of aerial RS compared to spaceborne
RS in terms of atmospheric disturbance?
2. How important are laboratory spectra in understanding the remote sensing images?

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

These are typical exam questions:
1. List the two models used to describe electromagnetic energy.
2. How are wavelength and frequency related to each other (give a formula)?
3. What is the electromagnetic spectrum?

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

4. List and define the three types of atmospheric scattering?

5. What specific energy interactions take place when EM energy from the Sun
hits the Earths surface?
6. In your own words give a definition of an atmospheric window.

first

last

back

exit

zoom

contents

index

about

2.4. Energy interactions with the Earths surface

7. Indicate True or False: Only the wavelength region outside the main absorption bands of the atmospheric gases can be used for remote sensing.
8. Indicate True or False: The amount of energy detected by a remote sensing
sensor is a function of how energy is partitioned between its source and
the materials with which it interacts on its way to the detector.

first

last

back

exit

zoom

contents

index

about

Chapter 3
Sensors and platforms

first

last

back

exit

zoom

contents

index

about

3.1. Introduction

3.1

Introduction

In Chapter 2, the underlying principle of remote sensing was explained. Depending on the surface characteristics electromagnetic energy from the Sun or
active sensor is reflected or energy may be emitted by the Earth itself. This energy is measured and recorded. The resulting data can be used to derive information about surface characteristics.
The measurements of electromagnetic energy are made by sensors that are
attached to a static or moving platform. Different types of sensors have been
developed for different applications (Section 3.2). Aircraft and satellites are generally used to carry one or more sensors (Section 3.3). General references with
respect to missions and sensors are [13, 14]. Please refer to ITCs Aerospace Sensor Database for a complete and up-to-date overview.
The sensor-platform combination determines the characteristics of the resulting image data. For example, when a particular sensor is operated from a higher
altitude the total area imaged is increased while the level of detail that can be
observed is reduced (Section 3.4). Based on your information needs and on time
and budgetary criteria, you can determine which image data are most appropriate (Section 3.5).

first

last

back

exit

zoom

contents

index

about

3.2. Sensors

3.2

Sensors

A sensor is a device that measures and records electromagnetic energy. Sensors

can be divided into two groups. Passive sensors depend on an external source
of energy, usually the Sun (although sometimes the Earth itself). The group of
passive sensors cover the electromagnetic spectrum in the range from less than
1 picometre (gamma rays) to over 1 metre (micro and radio waves). The oldest
and most common type of passive sensor is the (photographic) camera.
Active sensors have their own source of energy. Measurements by active sensors are more controlled because they do not depend upon the (varying) illumination conditions. Active sensors include the laser altimeter (using infrared
light) and radar. Figure 3.1 gives an overview of the types of the sensors that
are introduced in this section. The camera, the multispectral scanner and the
imaging radar are explained in more detail in Chapters 4, 5 and 6 respectively.
For more information about spaceborne remote sensing you may refer to ITCs
Aerospace Sensor Database.

first

last

back

exit

zoom

contents

index

about

3.2. Sensors

Visible
domain

Passive
sensors

gamma ray
spectrometer

Optical
domain

- multispectral scanner
- imaging spectrometer

Microwave
domain

thermal
scanner

passive microwave
radiometer

- aerial camera
- video camera

Active
sensors

radar
altimeter

laser
scanner

imaging
radar

Figure 3.1: Overview of

the sensors that are introduced in this chapter

wavelength

first

last

back

exit

zoom

contents

index

about

3.2. Sensors

3.2.1 Passive sensors

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Gamma-ray spectrometer
The gamma-ray spectrometer measures the amount of gamma rays emitted by
the upper soil or rock layers due to radioactive decay. The energy measured
in specific wavelength bands provides information on the abundance of (radio
isotopes that relate to) specific minerals. Therefore, the main application is found
in mineral exploration. Gamma rays have a very short wavelength on the order
of picometres (pm)). Because of large atmospheric absorption of these waves
this type of energy can only be measured up to a few hundred metres above the
Earths surface. Example data acquired by this sensor are given in Figure 7.1.

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Aerial camera
The camera system (lens and film) is mostly found in aircraft for aerial photography. Low orbiting satellites and NASA Space Shuttle missions also apply
conventional camera techniques. The film types used in the camera enable electromagnetic energy in the range between 400 nm and 900 nm to be recorded.
Aerial photographs are used in a wide range of applications. The rigid and
regular geometry of aerial photographs in combination with the possibility to
acquire stereo-photography has enabled the development of photogrammetric
procedures for obtaining precise 3D coordinates. Although aerial photos are
used in many applications, principal applications include medium and large
scale (topographic) mapping and cadastral mapping. Today, analogue photos
are often scanned to be stored and processed in digital systems. Various examples of aerial photos are shown in Chapter 4.

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Video camera
Video cameras are sometimes used to record image data. Most video sensors are
only sensitive to the visible colours, although a few are able to record the nearinfrared part of the spectrum (Figure 3.2). Until recently, only analogue video
cameras were available. Today, digital video cameras are increasingly available,
some of which are applied in remote sensing. Mostly, video images serve to
provide low cost image data for qualitative purposes, for example, to provide
additional visual information about an area captured with another sensor (e.g.,
laser scanner or radar).

Figure 3.2:
Analogue
false colour video image
of De Lopikerwaard (NL).
Courtesy of Syntoptics

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Multispectral scanner
The multispectral scanner is an instrument that mainly measures the reflected
sunlight in the optical domain. A scanner systematically scans the Earths
surface thereby measuring the energy reflected from the viewed area. This is
done simultaneously for several wavelength bands, hence the name multispectral scanner. A wavelength band is an interval of the electromagnetic spectrum
for which the average reflected energy is measured. The reason for measuring a
number of distinct wavelength bands is that each band is related to specific characteristics of the Earths surface. For example, reflection characteristics of blue
light give information about the mineral composition; reflection characteristics
of infrared light tell something about the type and health of vegetation. The
definition of the wavebands of a scanner, therefore, depends on the applications
for which the sensor has been designed. An example of multispectral data for

Figure 3.3:
Landsat
TM colour composite
(RGB=457) of Yemen.

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
geological applications is given in Figure 3.3.

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Imaging spectrometer
The principle of the imaging spectrometer is similar to that of the multispectral
scanner, except that spectrometers measure only very narrow (510 nm) spectral
bands. This results in an almost continuous reflectance curve per pixel rather
than the values for relatively broad spectral bands. The spectral curves measured depend on the chemical composition of the material. Imaging spectrometer data, therefore, can be used to determine mineral composition of the surface

30100+

TSM

Figure 3.4:
Total Suspended Matter concentration of the North Sea derived from SeaWIFS (OrbView2) data. Courtesy of
CCZM, Rijkswaterstaat

0
SeaWIFS
2-bands MIM
Atm correction: Modtran

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
or the chlorophyll content of the surface water (Figure 3.4).

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Thermal scanner
Thermal scanners measure thermal data in the range of 1014 m. Wavelengths
in this range are directly related to an objects temperature. Data on cloud, land
and sea surface temperature are extremely useful for weather forecasting. For
this reason, most remote sensing systems designed for meteorology include a
thermal scanner. Thermal scanners can also be used to study the effects of
drought (water stress) on agricultural crops, and to monitor the temperature
of cooling water discharged from thermal power plants. Another application is
in the detection of coal fires (Figure 3.5).

Figure 3.5: Night-time airborne thermal scanner image of a coal mining area.
Dark tones represent the
relatively cold surfaces,
whilst light tones represent
the relatively warm spots.
Most of the warm spots
are due to underground
coal fires apart from the
largest light patch which is
a lake.
first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Radiometer
EM energy with very long wavelengths (1100 cm) is emitted from the soil and
rocks on, or just below, the Earths surface. The depth from which this energy is
emitted depends on the properties, such as water content, of the specific material. Radiometers are used to detect this energy. The resulting data can be used
in mineral exploration, soil mapping and soil moisture estimation (Figure 3.6).

Figure 3.6: Map of the

Iberian Peninsula showing
the long-term trend in soil
moisture change (vol.%)
over a 10-year period.
Map is based on Nimbus/SMMR observations.
Courtesy of Free University Amsterdam / NASA

first

last

back

exit

zoom

contents

index

about

3.2. Sensors

3.2.2 Active sensors

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Laser scanner
Laser scanners are mounted on aircraft and use a laser beam (infrared light) to
measure the distance from the aircraft to points located on the ground. This
distance measurement is then combined with exact information on the aircrafts
position to calculate the terrain elevation. Laser scanning is mainly used to produce detailed, high-resolution, Digital Terrain Models (DTM) for topographic
mapping (Figure 3.7). Laser scanning is increasingly used for other purposes,
such as the production of detailed 3D models of city buildings and for measuring tree heights in forestry.

Figure 3.7: Digital Terrain

Model (5 m grid) of the St.
Pietersgroeve (NL). Courtesy Survey Department,
Rijkswaterstaat

first

last

back

exit

zoom

contents

index

about

3.2. Sensors
Radar altimeter
Radar altimeters are used to measure the topographic profile parallel to the satellite orbit. They provide profiles (single lines of measurements) rather than image data. Radar altimeters operate in the 16 cm domain and are able to determine height with a precision of 24 cm. Radar altimeters are useful for measuring relatively smooth surfaces such as oceans and for small scale mapping of
continental terrain models. Sample results of radar altimeter measurements are
given in Figure 1.6 and Figure 7.2.

first

last

back

exit

zoom

contents

index

about

100

3.2. Sensors
Imaging radar
Radar instruments operate in the 1100 cm domain. As in multispectral scanning, different wavelength bands are related to particular characteristics of the
Earths surface. The radar backscatter (Figure 3.8) is influenced by the illuminating signal (microwave parameters) and the illuminated surface characteristics
(orientation, roughness, di-electric constant/moisture content). Since radar is an
active sensor system and the applied wavelengths are able to penetrate clouds, it
has all-weather day-and-night acquisition capability. The combination of two
radar images of the same area can provide information about terrain heights.
Combining two radar images acquired at different moments can be used to precisely assess changes in height or vertical deformations (SAR Interferometry).

Figure 3.8: ERS SAR image of a delta in Kalimantan, Indonesia. The image allows three different
forest types to be distinguished.

first

last

back

exit

zoom

contents

index

about

101

3.3. Platforms

3.3

Platforms

In remote sensing, the sensor is mounted on a platform. In general, remote sensing sensors are attached to moving platforms such as aircraft and satellites. Static
platforms are occasionally used in an experimental context. For example, by using a multispectral sensor mounted to a pole, the changing reflectance characteristics of a specific crop during the day or season can be assessed.
Airborne observations are carried out using aircraft with specific modifications to carry sensors. An aircraft that carries an aerial camera or a scanner
needs a hole in the floor of the aircraft. Sometimes Ultra Light Vehicles (ULVs),
balloons, Airship or kites are used for airborne remote sensing. Airborne observations are possible from 100 m up to 3040 km height. Until recently, the
navigation of an aircraft was one of the most difficult and crucial parts of airborne remote sensing. In recent years, the availability of satellite navigation
technology has significantly improved the quality of flight execution.
For spaceborne remote sensing, satellites are used. Satellites are launched into
space with rockets. Satellites for Earth Observation are positioned in orbits between 15036,000 km altitude. The specific orbit depends on the objectives of
the mission, e.g., continuous observation of large areas or detailed observation
of smaller areas.

first

last

back

exit

zoom

contents

index

about

102

3.3. Platforms

3.3.1 Airborne remote sensing

Airborne remote sensing is carried out using different types of aircraft depending on the operational requirements and budget available. The speed of the
aircraft can vary between 140600 km/hour and is, among others, related to the
mounted sensor system. Apart from the altitude, also the aircraft orientation
affects the geometric characteristics of the remote sensing data acquired. The
orientation of the aircraft is influenced by wind conditions and can be corrected
for to some extent by the pilot. Three different aircraft rotations relative to a
reference path are possible: roll, pitch and yaw (Figure 3.9). An Inertial Measurement Unit (IMU) can be installed in the aircraft to measure these rotations.
Subsequently the measurements can be used to correct the sensor data for the
resulting geometric distortions.

+
0
-

roll angle

Figure 3.9: The three angles (roll, pitch and yaw)

of an aircraft that influence
image data acquired.

+
0
-

pitch angle

yaw angle

Today, most aircraft are equipped with satellite navigation technology, which
yield the approximate position (RMS-error of less than 30 m). More precise positioning and navigation (up to decimetre accuracy) is possible using so-called
differential approaches. In this textbook we refer to satellite navigation in general, which comprises the American GPS system, the Russian Glonass system
and the proposed European Galileo system.
In aerial photography the measurements are stored on hard-copy material:
the negative film. For other sensors, e.g., a scanner, the digital data can be stored
first

last

back

exit

zoom

contents

index

about

103

3.3. Platforms
on tape or mass memory devices. Tape recorders offer the fastest way to store
the vast amount of data. The recorded data are only available after the aircraft
has returned to its base.
Owning, operating and maintaining survey aircraft, as well as employing a
professional flight crew is an expensive undertaking. In the past, survey aircraft
were owned mainly by large national survey organizations that required large
amounts of photography. There is an increasing trend towards contracting specialized private aerial survey companies. Still, this requires basic understanding
of the process involved. A sample contract for outsourcing aerial photography
is provided by the American Society of Photogrammetry and Remote Sensing at
their ASPRS web site.

first

last

back

exit

zoom

contents

index

about

104

3.3. Platforms

3.3.2 Spaceborne remote sensing

Spaceborne remote sensing is carried out using sensors that are mounted on
satellites. The monitoring capabilities of the sensor are to a large extent determined by the parameters of the satellites orbit. Different types of orbits are required to achieve continuous monitoring (meteorology), global mapping (land
cover mapping), or selective imaging (urban areas). For remote sensing purposes, the following orbit characteristics are relevant :
altitude, which is the distance (in km) from the satellite to the mean surface
level of the Earth. Typically, remote sensing satellites orbit either at 600
800 km (polar orbit) or at 36,000 km (geo-stationary orbit) distance from
the Earth. The distance influences to a large extent which area is viewed
and at which detail.
inclination angle, which is the angle (in degrees) between the orbit and the
equator. The inclination angle of the orbit determines, together with field
of view of the sensor, which latitudes can be observed. If the inclination
is 60 then the satellite flies over the Earth between the latitudes 60 South
and 60 North; it cannot observe parts of the Earth at latitudes above 60 .
period, which is the time (in minutes) required to complete one full orbit.
A polar satellite orbits at 800 km altitude and has a period of 90 minutes.
A ground speed of 28,000 km/hour is almost 8 km/s. Compare this figure
with the speed of an aircraft, which is around 400 km/hour. The speed of
the platform has implications for the type of images that can be acquired
(time for exposure).
repeat cycle, which is the time (in days) between two successive identical
orbits. The revisit time, the time between two subsequent images of the
first

last

back

exit

zoom

contents

index

about

105

3.3. Platforms
same area, is determined by the repeat cycle together with the pointing
capability of the sensor. Pointing capability refers to the possibility of the
sensor-platform to look sideways. Pushbroom scanners, such as those
mounted on SPOT, IRS and IKONOS (Section 5.4), have this possibility.
The following orbit types are most common for remote sensing missions:
Polar, or near polar, orbit. These are orbits with inclination angle between
80 and 100 degrees and enable observation of the whole globe. The satellite
is typically placed in orbit at 600800 km altitude.
Sun-synchronous orbit. An orbit chosen in such a way that the satellite always
passes overhead at the same local solar time is called sun-synchronous. Most
sun-synchronous orbits cross the equator at mid-morning (around 10:30 h). At
that moment the Sun angle is low and the resultant shadows reveal terrain relief. Sun-synchronous orbits allow a satellite to record images at two fixed times
during one 24-hour period: one during the day and one at night. Examples of
near polar sun-synchronous satellites are Landsat, SPOT and IRS.
Geostationary orbit. This refers to orbits in which the satellite is placed above
the equator (inclination angle is 0 ) at a distance of some 36,000 km. At this
distance, the period of the satellite is equal to the period of the Earth. The result
is that the satellite is at a fixed position relative to the Earth. Geostationary orbits
are used for meteorological and telecommunication satellites.
Todays meteorological weather satellite systems use a combination of geostationary satellites and polar orbiters. The geo-stationary satellites offer a continuous view, while the polar orbiters offer a higher resolution (Figure 3.10).
The data of spaceborne sensors need to be sent to the ground for further analysis and processing. Some older spaceborne systems utilized film cartridges that
fell back to a designated area on Earth. In the meantime, practically all Earth Observation satellites apply satellite communication technology for downlink of the
first

last

back

exit

zoom

contents

index

about

106

3.3. Platforms

Figure 3.10:
Meteorological
observation
system
comprised
of
geo-stationary and polar
satellites.

data. The acquired data are sent down to a receiving station or to another communication satellite that downlink the data to receiving antenn on the ground.
If the satellite is outside the range of a receiving station the data can be temporarily stored by a tape recorder in the satellite and transmitted later. One of
the trends is that small receiving units (consisting of a small dish with a PC) are
being developed for local reception of image data.

first

last

back

exit

zoom

contents

index

about

107

3.4. Image data characteristics

3.4

Image data characteristics

Remote sensing image data are more than a picturethey are measurements of
EM energy. Image data are stored in a regular grid format (rows and columns).
The single elements are called pixels, an abbreviation of picture elements. For
each pixel, the measurements are stored as Digital Number values or DN-values.
Typically, for each wavelength band measured, a separate layer is stored (Figure 3.11).
Columns

Rows
45

band 3

26 81
53 35 57

Figure 3.11: An image file

comprises a number of
bands. For each band the
Digital Number (DN) values, corresponding to the
measurements, are stored
in a row-column system.

band 2
band 1

DN-values

Single pixel

The quality of image data is primarily determined by characteristics of the

sensor-platform system. These sensor-platform characteristics are usually referred to as:
1. Spectral resolution and radiometric resolution, referring respectively to the
part of the Electro Magnetic spectrum measured and the differences in energy that can be observed.
first

last

back

exit

zoom

contents

index

about

108

3.4. Image data characteristics

2. Spatial resolution, referring to the smallest unit-area measured, it indicates
the minimum size of objects that can be detected.
3. Revisit time, the time between two successive image acquisitions over the
same location on Earth.
The resolutions need a more complicated, sensor specific, explanation. These
are explained in the respective chapters (camera, multispectral scanner, radar).
The image data produced by the sensor-platform system are related to the above
mentioned resolutions. Characteristics of image data are:
1. Image size: the number of rows and columns in a scene.
2. Number of bands: the number of wavelengths band stored. For example, 1
(black/white photograph), 4 (SPOT multispectral image), or 256 (imaging
spectroscopy data).
3. Quantization: the data format used to stored the energy measurements.
Typically, for each measurement one (8 bits) byte is used, which represents
the (discrete) values of 0255. These values are known as Digital Numbers. Using sensor specific calibration parameters, the DN-values can be
converted into measured energy (Watt).
4. Ground pixel size: the area coverage of a pixel on the ground. Usually this
is a round figure (e.g., 20 m or 30 m). The ground pixel size of image data
is related to, but not necessarily the same as the spatial resolution.
The image size, the number of bands and the quantification, allow the disk
space required to be calculated. For example, a SPOT multispectral image requires: 3000 (columns) 3000 (rows) 4 (bands) 1 byte = 36 Mbyte of storage.
first

last

back

exit

zoom

contents

index

about

109

3.5. Data selection criteria

3.5

4. Which aspects need to be considered to assess if the statement RS data

acquisition is a cost-effective method is to be true?

first

last

back

exit

zoom

contents

index

about

116

3.5. Data selection criteria

The following are sample exam questions:
1. Explain the sensor-platform concept.
2. Mention two types of passive and two types of active sensor.
3. What is a typical application of a multispectral satellite image, and, what
is a typical application of a very high spatial resolution satellite image?

first

last

back

exit

zoom

contents

index

about

117

3.5. Data selection criteria

4. Describe two differences between aircraft and satellite remote sensing and
their implications for the data acquired.
5. Which two types of satellite orbits are mainly used for Earth observation?
6. List and describe four characteristics of image data.

first

last

back

exit

zoom

contents

index

about

Chapter 4
Aerial cameras

first

last

back

exit

zoom

contents

index

about

118

119

4.1. Introduction

4.1

Introduction

Aerial photography has been used since the early 20th century to provide spatial
data for a wide range of applications. It is the oldest, yet most commonly applied remote sensing technique. Photogrammetry is the science and technique
of making measurements from photos or image data. Nowadays, almost all topographic maps are based on aerial photographs. Aerial photographs also provide the accurate data required for many cadastral surveys and civil engineering projects. Aerial photography is a useful source of information for specialists
such as foresters, geologists and urban planners. General references for aerial
photography are [10, 16, 17].
Two broad categories of aerial photography can be distinguished: vertical
photography and oblique photography (Figure 4.1). In most mapping applications,
vertical aerial photography is required. Vertical aerial photography is produced
with a camera mounted in the floor of an aircraft. The resulting image is rather
similar to a map and has a scale that is approximately constant throughout the
image area. Usually, vertical aerial photography is also taken in stereo, in which
successive photos have a degree of overlap to enable stereo-interpretation and
stereo-measurements.
Oblique photographs are obtained when the axis of the camera is not vertical. Oblique photographs can be made using a hand-held camera and shooting through the (open) window of an aircraft. The scale of an oblique photo
varies from the foreground to the background. This scale variation complicates
the measurement of positions from the image and, for this reason, oblique photographs are rarely used for mapping purposes. Nevertheless, oblique images
can be useful for purposes such as viewing sides of buildings and for inventories
of wildlife.

first

last

back

exit

zoom

contents

index

about

120

4.1. Introduction

(a)

nadir

Figure 4.1: Vertical (a)

and oblique (b) photography

(b)

This chapter focusses on the camera, films and methods used for vertical
aerial photography. First of all, Section 4.2 introduces the aerial camera and its
main components. Photography is based on exposure of a film, processing and
printing. The type of film applied largely determines the spectral and radiometric characteristics of the printed products (Section 4.3). Section 4.4 focusses on
the geometric characteristics of aerial photography. In Section 4.5 some aspects
of aerial photography missions are introduced. In the advanced Section 4.6,
some technological developments are discussed.

first

last

back

exit

zoom

contents

index

about

121

4.1. Introduction

(a)

first

(b)

last

back

exit

zoom

contents

index

about

Figure 4.2: Vertical (a)

and oblique (b) aerial
photo of the ITC building.
Photos by Paul Hofstee,
1999.

122

4.2. Aerial camera

4.2

Aerial camera

A camera used for vertical aerial photography for mapping purposes is called an
aerial survey camera. In this section the standard aerial camera is introduced.
At present, there are only two major manufacturers of aerial survey cameras,
namely Zeiss and Leica. These two companies produce the RMK-TOP and the
RC-30 respectively. Just like a typical hand-held camera, the aerial survey camera contains a number of common components as well as a number of specialized ones necessary for its specific role. Figure 4.3 shows a schematic drawing
of an aerial camera. The large size of the camera results from the need to acquire
images of large areas with a high spatial resolution. This is realized by using
a large film size. Modern aerial survey cameras produce negatives measuring
23 cm 23 cm. Up to 600 photographs may be recorded on a single roll of film.

Film Magazine
Intervalometer
Camera Body
Camera

Mount
Viewfinder
Lens Cone

Exposure Control
System

first

last

back

exit

zoom

contents

Figure 4.3: Major components of an aerial camera

index

about

123

4.2. Aerial camera

4.2.1 Lens cone

Perhaps the most important (and expensive) single component within the camera is the lens cone. This is interchangeable, and the manufactures produce a
range of cones, each of different focal length. Focal length is the most important property of a lens cone since, together with flying height, it determines the
photo scale (Section 4.4.1). The focal length also determines the angle of view of
the camera. The longer the focal length, the narrower the angle of view. Lenses
are usually available in the following standard focal lengths, ranging from narrow angle (610 mm), to normal angle (305 mm) to wide angle (210 mm, 152 mm)
to super-wide angle (88 mm). The 152 mm lens is the most commonly used lens.
The lens cone is responsible for projecting an optical image onto the film.
The accuracy of this projection depends on the quality of the lens. Even with
high quality lenses some distortions still take place. These distortions are imperceptible to the human eye, but aversely affect the photogrammetric operations
where very precise measurements (at m level) are required. Therefore, the distortion of lenses is measured on a regular basis and reported in a calibration
report. This report is required in many photogrammetric processes.
In the lens cone (Figure 4.4), the diaphragm and shutter respectively control
the intensity and the duration of the exposure reaching the film. Typically, optical
filters are placed over the lens to control two important aspects of the image
quality.
Image contrast. A yellow filter is often used (in black and white photography) to absorb the ultra-violet and blue wavelengths, which are the most
highly scattered within the atmosphere (Section 2.3.2). This scattering effect, if not corrected, normally leads to a reduction in image contrast.

first

last

back

exit

zoom

contents

index

about

CAUTION

124

4.2. Aerial camera

Focal Plane
Frame

Optical Axis

Lens
System
Diaphragm

Figure 4.4: The lens cone

comprises a lens system,
focal plane frame, shutter,
diaphragm, anti-vignetting
and coloured filters.

Shutter

Coloured +
A.V. Filter

Evenness of illumination. Any image formed by a lens is brightest at the

centre and darkest in the corners. This is known as light fall-off or vignetting, and is related to the angle a light ray makes with the optical axis,
the greater the angle the darker the image. In wide angle lenses, this effect can be severe and images appear dark in the corners. The effect can
be partially corrected by an anti-vignetting filter, which is a glass plate in
which the centre transmits less light than the corners. In this way the image illumination is made more even over the whole image area, and the
dark corners are avoided.

first

last

back

exit

zoom

contents

index

about

125

4.2. Aerial camera

4.2.2 Film magazine and auxiliary data

The aerial camera is fitted with a system to record various items relevant information onto the side of the negative: mission identifier, date and time, flying
height and the frame number (Figure 4.5).
Altimeter

Enschede
25 June
1991

NAGII 7142 213.40

Watch
Message
Pad

Lens Number &

Focal Length

Fiducial Marks

Frame
Number

Spirit
Level

Figure 4.5: Auxiliary data

annotation on an aerial
photograph

2205

A vacuum plate is used for flattening the film at the instant of exposure. Socalled fiducial marks are recorded in all corners of the film. The fiducial marks are
first

last

back

exit

zoom

contents

index

about

126

4.2. Aerial camera

required to determine the optical centre of the photo needed to align photos for
stereoviewing. The fiducials are also used to record the precise position of the
film in relation to the optical system, which is required in photogrammetric processes (interior orientation, Section 9.4.3). Modern cameras also have an image
motion compensation facility (Section 4.6).

first

last

back

exit

zoom

contents

index

about

127

4.2. Aerial camera

4.2.3 Camera mounting

The camera mounting enables the camera to be levelled by the operator. This is
usually performed at the start of a mission once the aircraft is in a stable configuration. Formal specifications for aerial survey photography usually require that
the majority of images are maintained within a few degrees (less than 3 ) of true
vertical. Errors exceeding this cause difficulties in photogrammetric processing
(Section 9.4.3).

first

last

back

exit

zoom

contents

index

about

CAUTION

128

4.3. Spectral and radiometric characteristics

4.3

Spectral and radiometric characteristics

Photographic recording is a multi-stage process that involves film exposure and

chemical processing (development). It is usually followed by printing.
Photographic film comprises a light sensitive emulsion layer coated onto a
base material (Figure 4.6). The emulsion layer contains silver halide crystals, or
grains, suspended in gelatine. The emulsion is supported on a stable polyester
base. Light changes the silver halide into silver metal, which, after processing
of the film, appears black. The exposed film, before processing, contains a latent
image.
Supercoat
Emulsion Layer
Subbing

Figure 4.6: Section of a

photographic film showing
the emulsion layer with the
silver halide crystals

Base Layer
Anti-Halation Layer

The film emulsion type applied determines the spectral and radiometric characteristics of the photograph. Two terms are important in this context:
Spectral sensitivity describes the range of wavelengths to which the emulsion is sensitive. For the study of vegetation the near-infrared wavelengths
yield much information and should be recorded; for other purposes a standard colour photograph normally can yield the optimal basis for interpretation.

first

last

back

exit

zoom

contents

index

about

129

4.3. Spectral and radiometric characteristics

General sensitivity is a measure of how much light energy is required to
bring about a certain change in film density. Given specific illumination
conditions, the general sensitivity of a film can be selected, for example, to
minimize exposure time.
For reasons of understanding, first the general sensitivity is explained, followed by an explanation of the spectral sensitivity. Subsequently black-white,
colour and colour infrared photography are explained. The last section presents
some remarks about the scanning of photos.

first

last

back

exit

zoom

contents

index

about

130

4.3. Spectral and radiometric characteristics

4.3.1 General sensitivity

The energy of a light photon is inversely proportional to the light wavelength. In
the visible range, therefore, the blue light has the highest energy. For a normal
silver halide grain, only blue light photons have sufficient energy to form the
latent image, and hence a raw emulsion is only sensitive to blue light.
The sensitivity of a film can be increased by increasing the mean size of the
silver grains: larger grains produce more metallic silver per input light photon.
The mean grain size of aerial films is in the order of a few m. There is a problem
related to increasing the grain size: larger grains are unable to record small details, i.e., the spatial resolution is decreased (Section 4.4.2). The other technique
to improve the general sensitivity of a film is to perform a sensitization of the
emulsion by adding small quantities of chemicals, such as gold or sulphur.
General sensitivity is often referred to as film speed. For a scene of given average brightness, the higher the film speed, the shorter the exposure time required
to record the optical image on the film. Similarly, the higher the film speed, the
less bright an object needs to be in order to be recorded upon the film.

first

last

back

exit

zoom

contents

index

about

131

4.3. Spectral and radiometric characteristics

4.3.2 Spectral sensitivity

1.0

0.8

Relative Sensitivity

Sensitization techniques are used not only to increase the general sensitivity but
also to produce films that are sensitive to longer wavelengths. By adding sensitizing dyes to the basic silver halide emulsion, the energy of longer light wavelengths becomes sufficient to produce latent images. In this way a monochrome
film can be made sensitive to green, red or infrared wavelengths (Section 3.3.2).
A blackandwhite (monochrome) type of film has one emulsion layer. Using sensitization techniques, different types of monochrome films are available.
Most common are the panchromatic and infra-red sensitive film. The sensitivity
curves of these films are shown in Figure 4.7.

0.6
0.4
0.2

0.4
0.2
Blue + Green + Red + Infra Red Sensitive

0.0

(a)

Figure 4.7:
Spectral
sensitivity curves of a
panchromatic film (a) and
a black/white infrared film
(b). Note the difference in
scaling on the x-axis

0.6

Blue, Green & Red Sensitive

0.0

400

500
600
Wavelength (nm)

700

400

500

600
700
Wavelength (nm)

800

900

(b)

Colour photography uses an emulsion with three sensitive layers to record

three wavelength bands corresponding to the three primary colours of the spectrum, i.e., blue, green and red. There are two types of colour photography: true
colour and false colour infrared.

first

last

back

exit

zoom

contents

index

CAUTION

about

132

4.3. Spectral and radiometric characteristics

4.3.3 Monochrome photography

Photography is a multi-stage process involving exposure, chemical processing
and, finally, printing.
During exposure in the camera an optical image of the original scene is projected by the lens onto the film. This image is effectively a two-dimensional pattern of light with varying intensity. The illumination of the film causes photons
to be absorbed by the crystals of the emulsion. This energy absorption causes a
number of the silver ions close to the surface of the crystal to be converted into
atoms of silver metal. The greater the number of absorbed photons, the greater
the number of silver atoms that are created. However, in the brief period of the
exposure, insufficient atoms of silver are created to cause a visible change in the
appearance of the crystals. This process is known as the formation of the latent
image.
When the film is later developed, in crystals containing the latent image, a
much greater quantity of metallic silver is formed by the development process.
Consequently, these areas of the film appear much darker than those where no
exposure was received. Undeveloped silver halide is next dissolved by a fixing
solution, and is then washed out of the emulsion layer. The resulting photographic image is referred to as a negative because bright areas in the terrain
appear dark on the film and vice versa.
In order to obtain a positive image, the negative must be photographically
printed. This may be done by either a contact process or by optical enlarging.
The photographic printing material contains a similar emulsion layer to the film,
but this is usually coated onto a paper base. In the printing process, the negative
is illuminated by a light source and, depending upon the amount of silver metal
in each area, either more or less light is allowed to pass through. In the dark areas
of the negative, (corresponding to bright parts of the subject), little light passes
first

last

back

exit

zoom

contents

index

about

CAUTION

133

4.3. Spectral and radiometric characteristics

Printing Exposure
Negative

Dense part
of Negative
transmits
little light

Clear part of
Negative
transmits
much light

Figure 4.8: Negative film

and positive print. Bright
areas (white) on the negative result in dark areas
(green) on the positive.

Positive Print

and forms only a small amount of silver in the positive image. Conversely, in
clear areas of the negative, much light passes and forms large quantities of silver
in the print (see Figure 4.8).

first

last

back

exit

zoom

contents

index

about

134

4.3. Spectral and radiometric characteristics

4.3.4 True colour and colour infrared photography

Colour photography uses the subtractive principle of colours, which is explained
in Section 10.2.2. According to this principle, an original colour is reproduced
by combinations of the three secondary colours: yellow, magenta and cyan.
A true colour film is made sensitive to the primary colours by coating three
separate colour sensitive layers (blue, green, red) on a common base. Each layer
thus records only one of the primary colours and produces quantities of complimentary coloured dye. The upper layer is sensitive to blue light only and contains a yellow dye forming coupler. A yellow filter layer absorbs any remaining
blue light and prevents it from exposing lower layers. The middle layer is sensitive to green light and contains a magenta dye forming coupler. Finally, the
lowest layer is sensitive to red and contains a cyan dye forming coupler (Figure 4.9).
Blue
Green
exposure exposure

CAUTION

Red
exposure

Blue sensitive layer

Yellow filter
Green sensitive layer
Red sensitive layer
Film base
Silver Halide Grain

Figure 4.9: Emulsion layers of a true colour film

Colour Coupler

first

last

back

exit

zoom

contents

index

about

135

4.3. Spectral and radiometric characteristics

In the development process, the developing agent reduces the exposed silver
halide to silver metal and the coloured dyes are produced. During development,
subtractive dyes are produced in each of the three layers. These layers produce
a quantity of the complimentary coloured dye in each layer that is proportional
to the brightness of the original primary colour. This means that exposure of a
high amount of green light produces much magenta dye in the film.
Colour negative printing uses paper that has a similar emulsion structure to
that of colour negative film, and is processed in a similar manner. For example,
a red colour in the scene produces cyan dye in the negative. The cyan colour in
the negative, when illuminated with white light, absorbs red light but transmits
blue and green light. This results in the formation of magenta and yellow dyes
in the print, which then appears red. Bleaching and fixing are performed to
remove all metallic silver and any remaining silver halide (Figure 4.10).
Blue
subject

Green
subject

Red
subject

Yellow
dye

Magenta
dye

Cyan
dye

Blue
image

first

last

Green
image

back

exit

Figure 4.10: True world,

colour negative and corresponding colour positive
image

Red
image

zoom

contents

index

about

136

4.3. Spectral and radiometric characteristics

The film used in a False Colour Infrared film comprises three layers: a green
sensitive layer containing a yellow dye forming coupler; a red sensitive layer
containing a magenta dye forming coupler; a near-infrared sensitive layer containing a cyan dye forming coupler. False colour infrared film, is sensitive to
the near-infrared wavelengths and is therefore particularly useful for recording
information about vegetation. While true colour is used mostly for topographic
and urban mapping, false colour photography is mostly used for mapping of
(semi-) natural vegetation and forest types.

Spatial characteristics

Two important properties of an aerial photograph are scale and spatial resolution. These properties are determined by sensor (lens cone and film) and platform (flying height) characteristics. Lens cones are produced with different focal
length. Focal length is the most important property of a lens cone since, together
with flying height, it determines the photo scale. The focal length also determines the angle of view of the camera. The longer the focal length, the narrower
the angle of view. The 152 mm lens is the most commonly used lens.

first

last

back

exit

zoom

contents

index

about

140

4.4. Spatial characteristics

4.4.1 Scale
The relationship between the photo scale factor, s, flying height, H, and lens focal
length, f, is given by
s=

H
.
f

(4.1)

Hence, the same scale can be achieved with different combinations of focal
length and flying height. If the focal length of a lens is decreased whilst the
flying height remains constant, then (also refer to Figure 4.12):
The image scale factor will increase and the size of the individual details in
the image becomes smaller. In the example shown in Figure 4.12, using a
150 mm and 300 mm lens at H=2000 m results in a scale factor of 13,333
and 6,666 respectively;
f=150mm

f=300mm

H=2000 m

74
41

3066 m

(a)

first

last

1533 m

back

exit

zoom

Figure 4.12: Effects of using a different focal length

(a: 150 mm, b: 300 mm)
when operating a camera
from the same height

(b)

contents

index

about

141

4.4. Spatial characteristics

The ground coverage increases. A 23 cm negative covers a length (and width)
of respectively 3066 m and 1533 m using a 150 mm and 300 mm lens. This
has implications on the number of photos required for the mapping of a
certain area, which, in turn, affects the subsequent processing (in terms of
labour) of the photos.
The angular field of view increases and the image perspective changes. The
total field of view in situations A and B respectively is 74 degrees and 41 degrees. When wide-angle photography is used for mapping, the measurement of height information (z dimension) in a stereoscopic model is more
accurate than when long focal length lenses are used. The combination of
a low flying height with a wide-angle lens can be problematic when there
are large terrain height differences or high man-made objects in the scene.
Some areas may become hidden behind taller objects. This phenomenon
is called the dead ground effect.

first

last

back

exit

zoom

contents

index

about

142

4.4. Spatial characteristics

4.4.2 Spatial resolution

While scale is a generally understood and applied term, the use of spatial resolution in aerial photography is quite difficult. Spatial resolution refers to the
ability to record small adjacent objects in an image. The spatial resolution of
monochrome aerial photographs ranges from 40 to 800 line pairs per mm. The
better the resolution of a recording system, the more easily the structure of objects on the ground can be viewed in the image. The spatial resolution of an
aerial photograph depends on:
the image scale factor: spatial resolution decreases as the scale factor increases;
the quality of the optical system: expensive high quality aerial lenses give
much better performance than the inexpensive lenses on amateur cameras;
the grain structure of the photographic filmthe larger the grains, the
poorer the resolution;
the contrast of the original objectsthe higher the target contrast, the better the resolution;
atmospheric scattering effectsthis leads to loss of contrast and resolution;
image motionthe relative motion between the camera and the ground
causes blurring and loss of resolution.
From the above list it can be concluded that the physical value of resolution
in an aerial photograph depends on a number of factors. The most variable
factor is the atmospheric condition, which can change from mission to mission,
and, during a mission.
first

last

back

exit

zoom

contents

index

about

143

4.5. Aerial photography missions

4.5

first

Aerial photography missions

last

back

exit

zoom

contents

index

about

144

4.5. Aerial photography missions

Mission planning
When a mapping project requires aerial photographs, one of the first tasks is
to select the required photo scale factor, the type of lens to be used, the type
of film to be used and the required percentage of overlap. Forward overlap
usually is around 60%; sideways overlap typically is around 20%. Figure 4.13
shows a survey area that is covered by a number of flight lines. The date and
time of acquisition should be considered with respect to growing season, light
conditions and shadowing effects.
Forward overlap along flight-line

Sideways
overlap
between
flight lines

Survey Area
3

Figure 4.13: Example survey area for aerial photography. Note the forward
and sideways overlap of
the photographs.

4.6

Advances in aerial photography

The most significant improvements to standard aerial photography made during the last decade can be summarized as follows:
Global Navigation Satellite Systems (GPS-USA, Glonass-SU and proposed
Galileo-EU) provide a means of achieving accurate navigation. They offer precise positioning of the aircraft along the survey run, ensuring that
the photographs are taken at the correct points. This method of navigation is especially important in survey areas where topographic maps do
not exist, are old, are of small scale or of poor quality. It is also helpful
in areas where the terrain has few features (deserts, forests etc) because in
these cases conventional visual navigation is particularly difficult. The major aerial camera manufacturers (as well as some independent suppliers),
now offer complete software packages that enable the flight crew to plan,
execute and evaluate an entire aerial survey mission. Before the mission,
the boundaries of the survey area to be photographed are first entered into
the software (either from a keyboard or digitizing table), along with basic
mission parameters such as the required scale, lens focal length, overlap
requirements, terrain elevation et cetera. The program then calculates the
optimum positions of the required flight lines as well as the positions of the
individual photo-centres. This pre-calculated information is then stored
on a diskette and taken into the aircraft in order to execute the mission.
During the flight, the crew is presented with a moving map display showing the position of the aircraft, the required flight lines and the individual photo-centres. Either flying the aircraft manually or on auto-pilot, the
display is followed and the software instructs the camera precisely where
to take the photographs. After the mission the co-ordinates of the phofirst

last

back

exit

zoom

contents

index

about

CAUTION

148

first

last

back

exit

zoom

contents

index

about

154

155

5.1. Introduction

5.1

Introduction

Multispectral scanners measure reflected electromagnetic energy by scanning

the Earths surface. This results in digital image data, of which the elementary
unit is a picture element: pixel. As the name multispectral suggests, the measurements are made for different ranges of the EM spectrum. Multispectral scanners
have been used in remote sensing since 1972 when the first Landsat satellite was
launched. After the aerial camera it is the most commonly used sensor. Applications of multispectral scanner data are mainly in the mapping of land cover,
vegetation, surface mineralogy and surface water.
Two types of multispectral scanners are distinguished: the whiskbroom scanner and the pushbroom scanner. The principles of these scanners and their characteristics are explained in Section 5.2 and Section 5.3 respectively. Multispectral scanners are mounted on airborne and spaceborne platforms. Section 5.4
describes the most widely used satellite based scanners.

first

last

back

exit

zoom

contents

index

about

156

5.2. Whiskbroom scanner

5.2

Whiskbroom scanner

A combination of a single detector plus a rotating mirror can be arranged in

such a way that the detector beam sweeps in a straight line over the Earth
across the track of the satellite at each rotation of the mirror. In this way, the
Earths surface is scanned systematically line by line as the satellite moves forward. Because of this sweeping motion, the whiskbroom scanner is also known as
the across-track scanner (Figure 5.1). The first multispectral scanners applied the
whiskbroom principle. Today, many scanners are still based on this principle:
NOAA/AVHRR and Landsat/TM, for instance (Recall: platform/sensor).

5.2. Whiskbroom scanner

5.2.2 Geometric characteristics

At any instant the mirror of the whiskbroom scanner sees a circle-like area on
the ground. Directly below the platform (at nadir), the diameter, D, depends on
the viewing angle of the system, , and the height, H:
(5.2)

D = H.

D should be expressed in metres, in radians and H in metres. Consider a

scanner with = 2.5 mrad that is operated at 4000 m. Using Formula 5.2 one
can easily calculate that the diameter of the area observed under the platform is
10 m.
The viewing angle of the system () is also referred to as Instantaneous Field
of View, abbreviated as IFOV. The IFOV determines the spatial resolution of a
scanner. The Field of View (FOV) describes the total angle that is scanned. For
aircraft scanners it is usually expressed as an angle; for satellite based scanners
with a fixed height the effective image width is used.
Consider a single line scanned by a whiskbroom scanner mounted to a static
platform. This results in a series of measurements going from the left to the right
side. The values for single pixels are calculated by integrating over a carefully
selected time interval.

first

last

back

5.3.1 Spectral characteristics

To a large extent, the characteristics of a solid state detector are also valid for
a CCD-array. In principle, one CCD-array corresponds to a spectral band and
all the detectors in the array are sensitive to a specific range of wavelengths.
With current state-of-the-art technology, CCD-array sensitivity stops at 2.5 m
wavelength. If longer wavelengths are to be measured, other detectors need to
be used.
One drawback of CDD arrays is that it is difficult to produce an array in
which all the elements have similar sensitivity. Differences between the detectors may be visible in the recorded images as vertical banding.

first

last

back

exit

zoom

contents

index

about

163

5.3. Pushbroom scanner

5.3.2 Geometric characteristics

back

exit

zoom

contents

index

about

170

5.4. Some operational spaceborne multispectral scanners

first

Band
1

Wavelength (m)
0.450.52

0.520.60

0.630.69

0.760.90

1.551.75

10.412.5

2.082.35

last

System
Orbit
Sensor
Swath width
Off-track viewing
Revisit time
Spectral bands (m)

Spatial resolution
Data archive at

first

Table 5.8:
teristics

Terra
705 km, 98.2 , sun-synchronous, 10:30 AM
crossing, 16 days repeat cycle
ASTER
60 km
Yes, 8.5 SWIR and TIR 24 VNIR
5 days
VNIR 0.0.56(1), 0.66(2), 0.81(3)
SWIR 1.65(1), 2.17(2), 2.21(3), 2.26(4),
2.33(5), 2.40(6)
TIR 8.3(1), 8.65(2), 9.10(3), 10.6(4),
11.3(5)
15 m (VNIR), 30 m(SWIR), 90 m(TIR)
terra.nasa.gov

last

back

exit

zoom

contents

index

about

Terra charac-

177

5.4. Some operational spaceborne multispectral scanners

5.4.8 EO-1
The EO-1 mission is part of the NASA New Millennium Program and is focused
on new sensor and spacecraft technologies that can directly reduce the cost of
Landsat and related Earth monitoring Systems. The EO-1 satellite is in an orbit
that covers the same ground track as Landsat 7, approximately one minute later.
This enables EO-1 to obtain images of the same ground area at nearly the same
time, so that direct comparison of results can be obtained from Landsat ETM+
and the three primary EO-1 instruments. The three primary instruments on the
EO-1 spacecraft are the Hyperion and the Linear Etalon Imaging Spectrometer
Array (LEISA), Atmospheric Corrector (LAC) and the Advanced Land Imager
(ALI).
Hyperion is a grating imaging spectrometer with a 30 m ground sample distance over a 7.5 km swath, providing 10 nm (sampling interval) contiguous
bands of the solar reflected spectrum from 4002500 nm. LAC is an imaging
spectrometer covering the spectral range from 9001600 nm which is well-suited
to monitor the atmospheric water absorption lines for correction of atmospheric
effects in multispectral imagers such as ETM+ on Landsat.
The Earth Observing-1 (EO-1) Advanced Land Imager (ALI) is a technology
verification instrument. Operating in a pushbroom fashion at an orbit of 705 km,
the ALI will provide Landsat-type panchromatic and multispectral bands. These
bands have been designed to mimic six Landsat bands with three additional
bands covering 0.4330.453, 0.8450.890, and 1.201.30 m. The ALI also contains wide-angle optics designed to provide a continuous 15 1.625 field of
view for a fully populated focal plane with 30 m resolution for the multispectral
pixels and 10 m resolution for the panchromatic pixels.

first

last

back

exit

zoom

contents

index

about

178

5.4. Some operational spaceborne multispectral scanners

System
Orbit
Sensor
Swath width
Off-track viewing
Revisit time
Spectral bands (m)
Spatial resolution
Data archive at

first

RADAR

first

last

back

exit

zoom

contents

index

about

184

185

6.1. What is radar?

6.1

What is radar?

So far, you have learned about remote sensing using the visible and infrared
part of the electromagnetic spectrum. Microwave remote sensing uses electromagnetic waves with wavelengths between 1 cm and 1 m (Figure 2.5). These
relatively longer wavelengths have the advantage that they can penetrate clouds
and are independent of atmospheric conditions, like haze. In microwave remote
sensing there are active and passive sensors. Passive sensors operate similarly
to thermal sensors by detecting naturally emitted microwave energy. They are
used in meteorology, hydrology and oceanography. In active systems, the antenna transmits microwave signals from an antenna to the Earths surface where
they are backscattered. The part of the electromagnetic energy that is scattered
into the direction of the antenna is detected by the sensor as illustrated in Figure 6.1. There are several advantages to be gained from the use of active sensors,
which have their own energy source:
It is possible to acquire data at any time including during the night (similar
to thermal remote sensing).
Since the waves are created actively, the signal characteristics are fully controlled (e.g., wavelength, polarization, incidence angle, et cetera) and can be
adjusted according to the desired application.
Active sensors are divided into two groups: imaging and non-imaging sensors.
RADAR sensors belong to the group of most commonly used active imaging microwave sensors. The term RADAR is an acronym for RAdio Detection And Ranging. Radio stands for the microwave and range is another term for distance. Radar
sensors were originally developed and used by the military. Nowadays, radar
first

last

back

exit

zoom

contents

index

about

186

6.1. What is radar?

Figure 6.1: Principle of

active microwave remote
sensing

sensors are widely used in civil applications too, such as environmental monitoring. To the group of non-imaging microwave instruments belong altimeters,
which collect distance information (e.g., sea surface height), and scatterometers,
which acquire information about the object properties (e.g., wind speed).
This chapter will focus on the principles of imaging radar and its applications. The interpretation of radar imagery is less intuitive than that obtained
from optical remote sensing. This is because of differences in physical interaction of the waves with the Earths surface. The chapter will explain which
interactions take place and how radar images can be interpreted.

first

last

back

exit

zoom

contents

index

about

187

6.2. Principles of imaging radar

6.2

Principles of imaging radar

Imaging radar systems include several components: a transmitter, a receiver,

an antenna and a recorder. The transmitter is used to generate the microwave
signal and transmit the energy to the antenna from where it is emitted towards
the Earths surface. The receiver accepts the backscattered signal as received by
the antenna, filters and amplifies it as required for recording. The recorder then
stores the received signal.
Imaging radar acquires an image in which each pixel contains a digital number according to the strength of the backscattered energy that is received from
the ground. The energy received from each transmitted radar pulse can be expressed in terms of the physical parameters and illumination geometry using
the so-called radar equation
Pr =

G 2 2 P t
,
(4)3 R4

(6.1)

where
Pr is the received energy,
G is the antenna gain,
is the wavelength,
Pt is the transmitted energy,
is the radar cross section, it is a function of the object characteristics and
the size of the illuminated area,

first

last

back

exit

zoom

contents

index

about

188

6.2. Principles of imaging radar

R is the range from the sensor to the object.
From this equation you can see that there are three main factors that influence
the strength of the backscattered received energy:
radar system properties, i.e., wavelength, antenna and transmitted power,
radar imaging geometry, that defines the size of the illuminated area which
is a function of i.e., beam-width,incidence angle and range,
object characteristics in relation to the radar signal, i.e., surface roughness
and composition, and terrain topography and orientation.
They are explained in the following sections in more detail.

first

last

back

exit

zoom

contents

index

about

189

6.2. Principles of imaging radar

What exactly does a radar system measure?
To interpret radar images correctly, it is important to understand what a radar
sensor detects. Imagine the transmitter creates microwave signals, i.e., pulses
of microwaves at regular intervals, the Pulse Repetition Frequency (PRF), that are
bundled by the antenna into a beam. This beam travels through the atmosphere,
illuminates a portion of the Earths surface, is backscattered and passes through
the atmosphere again to reach the antenna where the signal intensity is received.
From the time interval the signal needs to pass twice the distance between object and antenna, and knowing the speed of light, the distance (range) between
sensor and object can be derived.
To create an image, the return signal of a single pulse is sampled and these
samples are stored in an image line. With the movement of the sensor, emitting pulses, a two-dimensional image is created (each pulse defines one line).
The radar sensor, therefore, measures distances and detects backscattered signal
intensities.

first

last

back

exit

zoom

contents

index

about

190

6.2. Principles of imaging radar

Commonly used imaging radar bands
Similarly to optical remote sensing, radar sensors operate with different bands.
For better identification, a standard has been established that defines various
wavelength ranges using letters to distinguish among the various bands (Figure 6.2). In the description of different radar missions you will recognise the different wavelengths used if you see the letters. The European ERS mission and
the Canadian Radarsat, for example, use C-band radar. Just like multispectral
bands, different radar bands provide information about different object characteristics.
Band

first

Frequency (GHz)

0.3

1.0

3.0

10.0

30.0

100.0

Wavelength (cm)

100

0.3

last

back

exit

Q V

zoom

contents

Figure 6.2:
Microwave
spectrum
and
band
identification by letters

index

about

191

6.2. Principles of imaging radar

Microwave polarizations
The polarization of an electromagnetic wave is important in the field of radar
remote sensing. Depending on the orientation of the transmitted and received
radar wave, polarization will result in different images (Figure 6.3). It is possible
to work with horizontally, vertically or cross-polarized radar waves. Using different polarizations and wavelengths, you can collect information that is useful
for particular applications, for example, to classify agricultural fields. In radar
system descriptions you will come across the following abbreviations:
HH: horizontal transmission and horizontal reception,
VV: vertical transmission and vertical reception,
HV: horizontal transmission and vertical reception, and
VH: vertical transmission and horizontal reception.
Electric field

Wavelength, l

Figure 6.3: A vertically

polarized electromagnetic
wave; the electric fields
variation occurs in the vertical plane in this example.

Distance
Magnetic field

Velocity of light, c

first

last

back

exit

zoom

contents

index

about

192

6.3. Geometric properties of radar

6.3

Geometric properties of radar

The platform carrying the radar sensor moves along the orbit in flight direction (Figure 6.4). You can see the ground track of the orbit/flight path on the
Earths surface at nadir. The microwave beam illuminates an area, or swath, on
the Earths surface, with an offset from the nadir. The direction along-track is
called azimuth, the direction perpendicular (across-track) is called range.

SAR

Slant Range
Look angle

Groundtrack
Incidence Angle
Near Range
Mid Range

th
imu

Far Range

Radar Swath

Figure 6.4: Radar remote

sensing geometry

first

last

back

exit

zoom

contents

index

about

193

6.3. Geometric properties of radar

6.3.1 Radar viewing geometry

Radar sensors are side-looking instruments. The portion of the image that is
closest to the nadir track of the satellite carrying the radar is called near range.
The part of the image that is farthest from the nadir is called far range (Figure 6.4).
The incidence angle of the system is defined as the angle between the radar beam
and the local vertical. Moving from near range to far range, the incidence angle increases. It is important to distinguish between the incidence angle of the
sensor and the local incidence angle, which differs depending on terrain slope and
earth-curvature (Figure 6.5). It is defined as the angle between the radar beam
and the local surface normal. The radar sensor measures the distance between
antenna and object. This line is called slant range. But the true horizontal distance along the ground corresponding to each measure point in slant range is
called ground range (Figure 6.6).
Surface
normal

Vertical

Radar
wave

Local
slope

Local
incidence
angle

Incidence
angle

Figure 6.5: Radar incidence angle and local incidence angle

Scattering surface

first

last

back

exit

zoom

contents

index

about

194

6.3. Geometric properties of radar

sensor
1

slant range

altitude

Figure 6.6:
Geometric distortions in radar
imagery due to terrain
elevations

terrain
F
2
ground range 1
Foreshortening

first

last

L
3
4
Layover

6.3.3 Synthetic Aperture Radar (SAR)

Obviously there is a physical limit to the length of the antenna, the aperture
that can be carried on an aircraft or satellite. On the other hand, shortening the
wavelength has its limitations in penetrating clouds. Therefore, an approach
in which the aperture is increased synthetically is applied. Systems using this
approach are called Synthetic Aperture Radar (SAR). The synthesization of the
antenna length is achieved by taking advantage of the forward motion of the
platform and using several backscattered signals including the same object to
simulate a very long antenna. Most airborne and spaceborne radar systems use
this type of radar.

first

last

back

exit

zoom

contents

index

about

199

6.4. Distortions in radar images

6.4

Distortions in radar images

Due to the side-looking viewing geometry, radar images suffer from serious geometric and radiometric distortions. In radar imagery, you encounter variations
in scale (slant range to ground range conversion), foreshortening, layover and shadows (terrain elevation). Interference due to the coherency of the signal causes
speckle effects.

first

last

back

exit

zoom

contents

index

about

200

6.4. Distortions in radar images

6.4.1 Scale distortions

Radar measures ranges to objects in slant range rather than true horizontal distances along the ground. Therefore, the image has different scales moving from
near to far range (Figure 6.4). This means that objects in near range are compressed with respect to objects at far range. For proper interpretation, the image
has to be corrected and transformed into ground range geometry.

6.4.3 Radiometric distortions

The above-mentioned geometric distortions also have an influence on the received energy. Since the backscattered energy is collected in slant range the received energy coming from a slope facing the sensor is stored in a reduced area
in the image, i.e., it is compressed into fewer image pixels than should be the
case if obtained in ground range geometry. This results in high digital numbers
because the energy collected from different objects is combined. Unfortunately
this effect cannot be corrected. This is why especially layover and shadow areas
in radar imagery cannot be used for interpretation. However, they are useful
in the sense that they contribute to a three-dimensional look of the image and
therefore help the understanding of the terrain structure and topography.

(a)

(b)

A typical property of radar images is the so-called speckle. It appears as

grainy salt and pepper effects in the image (Figure 6.7). Speckle is caused by
the interaction of the different microwaves backscattered from the object area.
The wave interactions are called interference. Interference causes the return signals to be extinguished or amplified resulting in dark and bright pixels in the

first

6.5

Interpretation of radar images

The brightness of features in a radar image depends on the strength of the backscattered signal. In turn, the amount of energy that is backscattered depends on various factors. An understanding of these factors will help you to interpret radar
images properly.

6.5.2 Scattering patterns

The radar is scattered in different ways depending on the above-mentioned characteristics of the signal and object. Changes in the electrical properties influence
the absorption, transmission and reflection of microwaves. This means that the
moisture content of the surface is an essential component of the total scattering.
If an object is wet, surface scattering takes place. The type of reflection (ranging from specular to diffuse) and the strength depend on the roughness of the
material. Generally, reflectivity, and therefore image brightness, increases with
increasing moisture content.

first

last

back

exit

zoom

contents

index

about

214

6.5. Interpretation of radar images

6.5.3 Applications of radar

There are many useful applications of radar images. Radar data provide complementary information to visible and infrared remote sensing data. In the case
of forestry, radar images can be used to obtain information about forest canopy,
biomass and different forest types. Radar images also allow the differentiation of
different land cover types such as urban areas, agricultural fields, water bodies,
et cetera. In agricultural crop identification, the use of radar images acquired using different polarization (mainly airborne) is quite effective. It is crucial for agricultural applications to acquire data at a certain point in time (season) to obtain
the necessary parameters. This is possible because radar can operate independently of weather or daylight conditions. In geology and geomorphology the
fact that radar provides information about surface texture and roughness plays
an important role in lineament detection and geological mapping. Other successful applications of radar include hydrological modelling and soil moisture
estimation, based on the sensitivity of the microwave to the dielectric properties
of the observed surface. The interaction of microwaves with ocean surfaces and
ice provides useful data for oceanography and ice monitoring. Operational systems use data from the European SAR system ERS-2 and the Canadian Radarsat
programme. In this framework the data is also used for oil slick monitoring and
environmental protection. Looking at SAR interferometry there are plenty of interesting examples in the field of natural disaster monitoring and assessment,
i.e., earthquakes, volcano eruptions, flooding, et cetera.

first

last

back

exit

zoom

contents

index

about

215

6.6. Advanced radar processing techniques

6.6

Advanced radar processing techniques

Radar data provide a wealth of information that is not only based on a derived
intensity image but also on other data properties that measure characteristics of
the objects. One example is radar interferometry, an advanced processing method
that takes advantage of the phase information of the microwave. If you look at
two waves that are emitted with a slight offset you obtain a phase difference
between the two waves. The offset can be based on two antennas mounted
on the same platform (e.g., aircraft or space shuttle) or based on two different orbits/passes. The range difference is calculated by measuring the phase
difference between the two backscattered waves received at the antenna. With
the knowledge of the position of the platform with respect to the Earth surface
the elevation of the object is determined. Phase differences are displayed in socalled interferograms where different colours represent variations in height. The
interferogram is used to produce digital elevation models. Differential interferometry is based on the creation of two interferograms of successive radar data
acquisitions. These interferograms are subtracted from each other in order to
illustrate the changes that have occurred. These are useful for change detection,
e.g., earthquake damage assessment.

first

last

back

exit

zoom

contents

index

about

CAUTION

216

6.6. Advanced radar processing techniques

Summary
In this chapter, the principles of imaging radar and its applications have been
introduced. The microwave interactions with the surface have been explained
to illustrate how radar images are interpreted. Radar sensors measure distances
and detect backscattered signal intensities. In radar processing, special attention has to be paid to geometric corrections and speckle reduction for improved
interpretation. Radar data have many potential applications in the fields of geology, oceanography, hydrology, environmental monitoring, land use and land
cover mapping and change detection.

first

last

back

exit

zoom

contents

index

about

217

6.6. Advanced radar processing techniques

Questions
The following questions can help to study Chapter 6.
1. List three major differences between optical and microwave remote sensing?
2. What type of information can you extract from imaging radar data?
3. What are the limitations of radar images in terms of visual interpretation?

first

last

back

exit

zoom

contents

index

about

218

6.6. Advanced radar processing techniques

4. What kind of processing is necessary to prepare radar images for interpretation? Which steps are obligatory and which are optional?
5. Search the Internet for successful applications of radar images from ERS1/2, Radarsat and other sensors.

first

last

back

exit

zoom

contents

index

about

Chapter 7
Remote sensing below the ground
surface

first

last

back

exit

zoom

contents

index

CAUTION

about

219

220

7.1. Introduction

7.1

Introduction

The foregoing methods have relied on the electromagnetic spectrum in or near

the wavelength of visible light and are largely confined in their application to
the investigation of reflections and emissions from the Earths surface. To probe
more deeply into the ground, a range of methods that have their origin in the
physical or chemical properties of the buried rocks themselves may be employed.
These methods are often called geophysical methods. While the methods of
data collection may, of necessity, differ from the methods employed in others
types of remote sensing, the presentation of geophysical data nowadays uses
much of the technology originally developed for remote sensing data sensu stricto.
To maximize our ability to probe into the depth dimension, the integrated interpretation of all available data is to be recommended, so much is to be gained
by a closer integration of remote sensing and geophysics, while appreciating the
important physical differences in the origins of the images obtained. General
references to this subject are [9, 22].

first

last

back

exit

zoom

contents

index

about

221

7.2. Gamma-ray surveys

7.2

(a)

The following are typical exam questions:
1. Why is it necessary to use geophysical methods to explore the subsurface?
What are the limitations of visual RS methods in this respect?
2. Make a list of the physical properties of rocks that have been used as the
basis of geophysical mapping methods.
3. In the process of systematic exploration for Earth resources, why is it important to use inexpensive methods for the reconnaissance of large areas
before using more expensive methods over much smaller ones?

first

last

back

exit

zoom

contents

index

about

Chapter 8
Radiometric aspects

first

last

back

exit

zoom

contents

index

about

235

236

8.1. Introduction

8.1

Introduction

The previous chapters have examined remote sensing as a means of producing

image data for a variety of purposes. The following chapters deal with processing of the image data for rectification, visualization and interpretation. The first
step in the processing chain, often referred to as pre-processing, involves radiometric and geometric corrections (Figure 8.1). The radiometric aspects are dealt
with in this chapter and the geometric aspects in the following. Two groups of
radiometric corrections are identified:
the cosmetic rectification to compensate for data errors (Section 8.2), and
the atmospheric corrections to compensate for the effect of atmospheric
and illumination parameters, such as haze, sun angle and skylight on the
image data (Section 8.3).

Cosmetic
corrections

- Radiometric

corrections

- Atmospheric

Pre-processing

corrections
-

first

last

Geometric
corrections

back

exit

Figure 8.1: Image preprocessing steps

zoom

contents

index

about

237

8.2. Cosmetic corrections

8.2

Cosmetic corrections

Cosmetic corrections involve all those operations that are aimed at correcting
visible errors and noise in the image data. Defects in the data may be in the form
of periodic or random missing lines (line dropouts), line striping, and random
or spike noise. These effects can be identified visually and automatically.
34
28
23
17
13
11
10
10
10
10
15
25
33
39
43
46
48
50
51
52
53

27
21
17
14
12
11
11
11
11
11
18
31
40
44
47
48
50
52
53
54
55

20
16
14
13
12
11
11
11
11
13
21
36
45
47
49
50
52
54
55
56
57

17
15
14
12
11
11
11
11
13
17
26
40
47
49
50
51
53
55
57
58
59

17
17
15
13
11
11
11
13
17
25
33
42
48
50
51
53
55
57
58
58
58

19
18
16
12
11
11
12
16
23
33
40
46
49
51
52
54
56
58
59
59
58

20
18
15
11
10
12
16
22
30
40
47
49
51
53
54
55
56
58
59
59
58

21
18
15
12
12
15
20
28
36
45
50
50
51
53
54
54
55
56
57
57
57

22
20
17
14
16
22
28
36
42
47
50
50
50
51
52
52
53
54
55
55
55

21
19
16
13
15
23
31
39
45
48
50
50
50
50
51
51
51
51
52
52
52

19
16
13
11
13
21
29
39
45
48
50
50
50
50
51
51
50
50
49
49
48

18
14
12
12
16
24
32
42
48
50
51
51
52
52
51
51
50
50
49
49
49

16
14
14
17
23
31
39
47
51
53
54
54
54
54
53
53
52
52
51
51
52

17
17
19
25
31
39
46
52
55
55
55
55
55
55
55
55
55
55
54
54
55

22
24
28
34
40
46
51
55
57
57
57
57
57
57
57
57
57
58
59
59
61

(b)

(a)

first

last

back

exit

zoom

contents

index

about

Figure 8.2:
Simulated
Landsat MSS image of
a coastal area (a) and
the corresponding Digital
Numbers (DN) of a subset
of it (b).

238

8.2. Cosmetic corrections

8.2.1 Periodic line dropouts

Periodic line dropouts occur due to recording problems when one of the detectors of the sensor in question either gives wrong data or stops functioning. The
Landsat Thematic Mapper, for example, has 16 detectors in all its bands except
the thermal band. A loss of one of the detectors would result in every sixteenth
scan line being a string of zeros that would plot as a black line on the image. For
the Landsat MSS, this defect would occur in every sixth line (see Figure 8.3).
34
0
23
17
13
11
10
0
10
10
15
25
33
39
43
46
48
50
51
0
53

27
0
17
14
12
11
11
0
11
11
18
31
40
44
47
48
50
52
53
0
55

20
0
14
13
12
11
11
0
11
13
21
36
45
47
49
50
52
54
55
0
57

17
0
14
12
11
11
11
0
13
17
26
40
47
49
50
51
53
55
57
0
59

17
0
15
13
11
11
11
0
17
25
33
42
48
50
51
53
55
57
58
0
58

19
0
16
12
11
11
12
0
23
33
40
46
49
51
52
54
56
58
59
0
58

20
0
15
11
10
12
16
0
30
40
47
49
51
53
54
55
56
58
59
0
58

21
0
15
12
12
15
20
0
36
45
50
50
51
53
54
54
55
56
57
0
57

22
0
17
14
16
22
28
0
42
47
50
50
50
51
52
52
53
54
55
0
55

21
0
16
13
15
23
31
0
45
48
50
50
50
50
51
51
51
51
52
0
52

19
0
13
11
13
21
29
0
45
48
50
50
50
50
51
51
50
50
49
0
48

18
0
12
12
16
24
32
0
48
50
51
51
52
52
51
51
50
50
49
0
49

16
0
14
17
23
31
39
0
51
53
54
54
54
54
53
53
52
52
51
0
52

17
0
19
25
31
39
46
0
55
55
55
55
55
55
55
55
55
55
54
0
55

22
0
28
34
40
46
51
0
57
57
57
57
57
57
57
57
57
58
59
0
61

(a)

(b)

The first step in the restoration process is to calculate the average DN-value
per scan line for the entire scene. The average DN-value for each scan line is then
compared with this scene average. Any scan line deviating from the average by
more than a designated threshold value is identified as defective. In regions
of very diverse land cover, better results can be achieved by considering the
histogram for sub-scenes and processing these sub-scenes separately.
The next step is to replace the defective lines. For each pixel in a defective
first

last

back

exit

zoom

contents

index

about

Figure 8.3: The image

with line dropouts (a) and
the DN-values (b).

239

8.2. Cosmetic corrections

line, an average DN is calculated using DNs for the corresponding pixel in the
preceding and succeeding scan lines. The average DN is then substituted for
the defective pixel. The resulting image is a major improvement, although every
sixteenth scan line (or every sixth scan line, in case of Landsat MSS data) consists
of artificial data (see Figure 8.4). This restoration program is equally effective for
random line dropouts that do not follow a systematic pattern.
34
28
23
17
13
11
10
10
10
10
15
25
33
39
43
46
48
50
51
52

27
22
17
14
12
11
11
11
11
11
18
31
40
44
47
48
50
52
53
54

20
17
14
13
12
11
11
11
11
13
21
36
45
47
49
50
52
54
55
56

17
15
14
12
11
11
11
12
13
17
26
40
47
49
50
51
53
55
57
58

17
16
15
13
11
11
11
14
17
25
33
42
48
50
51
53
55
57
58
58

19
17
16
12
11
11
12
17
23
33
40
46
49
51
52
54
56
58
59
58

20
17
15
11
10
12
16
23
30
40
47
49
51
53
54
55
56
58
59
58

21
18
15
12
12
15
20
28
36
45
50
50
51
53
54
54
55
56
57
57

22
19
17
14
16
22
28
35
42
47
50
50
50
51
52
52
53
54
55
55

21
18
16
13
15
23
31
38
45
48
50
50
50
50
51
51
51
51
52
52

19
16
13
11
13
21
29
37
45
48
50
50
50
50
51
51
50
50
49
48

18
15
12
12
16
24
32
40
48
50
51
51
52
52
51
51
50
50
49
49

16
15
14
17
23
31
39
45
51
53
54
54
54
54
53
53
52
52
51
51

17
18
19
25
31
39
46
50
55
55
55
55
55
55
55
55
55
55
54
54

22
25
28
34
40
46
51
54
57
57
57
57
57
57
57
57
57
58
59
60

(b)

(a)

first

last

back

exit

zoom

contents

index

about

Figure 8.4: The image

after correction for line
dropouts (a) and the DNvalues (b).

240

8.2. Cosmetic corrections

8.2.2 Line striping

Line striping is far more common than line dropouts are. Line striping often
occurs due to non-identical detector response. Although the detectors for all
satellite sensors are carefully calibrated and matched before the launch of the
satellite, with time the response of some detectors may drift to higher or lower
levels. As a result, every scan line recorded by that detector is brighter or darker
than the other lines (see Figure 8.5). It is important to understand that valid
data are present in the defective lines, but these must be corrected to match the
overall scene.
34
96
23
17
13
11
10
20
10
10
15
25
33
78
43
46
48
50
51
89
53

27
92
17
14
12
11
11
22
11
11
18
31
40
88
47
48
50
52
53
91
55

20
92
14
13
12
11
11
22
11
13
21
36
45
93
49
50
52
54
55
99
57

17
90
14
12
11
11
11
22
13
17
26
40
47
95
50
51
53
55
57
99
59

17
94
15
13
11
11
11
26
17
25
33
42
48
95
51
53
55
57
58
99
58

19
88
16
12
11
11
12
32
23
33
40
46
49
97
52
54
56
58
59
99
58

20
86
15
11
10
12
16
44
30
40
47
49
51
98
54
55
56
58
59
99
58

21
86
15
12
12
15
20
56
36
45
50
50
51
98
54
54
55
56
57
97
57

22
94
17
14
16
22
28
72
42
47
50
50
50
96
52
52
53
54
55
96
55

21
98
16
13
15
23
31
78
45
48
50
50
50
94
51
51
51
51
52
94
52

19
92
13
11
13
21
29
78
45
48
50
50
50
94
51
51
50
50
49
98
48

18
88
12
12
16
24
32
84
48
50
51
51
52
96
51
51
50
50
49
98
49

16
87
14
17
23
31
39
94
51
53
54
54
54
98
53
53
52
52
51
96
52

17
84
19
25
31
39
46
95
55
55
55
55
55
99
55
55
55
55
54
94
55

22
88
28
34
40
46
51
96
57
57
57
57
57
99
57
57
57
58
59
94
61

(b)

(a)

Though several procedures can be adopted to correct this effect, the most
popular is the histogram matching. Separate histograms corresponding to each
detector unit are constructed and matched. Taking one response as standard, the
gain (rate of increase of DN) and offset (relative shift of mean) for all other detector units are suitably adjusted and new DN-values are computed and assigned.
first

last

back

exit

zoom

contents

index

about

Figure 8.5: The image

with line striping (a) and
the DN-values (b). Note
that the destriped image
would look almost similar
to the original image.

241

8.2. Cosmetic corrections

This yields a destriped image in which all DN-values conform to the reference
level and scale.

first

last

back

exit

zoom

contents

index

about

242

8.2. Cosmetic corrections

8.2.3 Random noise or spike noise

The periodic line dropouts and striping are forms of non-random noise that may
be recognised and restored by simple means. Random noise, on the other hand,
requires a more sophisticated restoration method such as digital filtering.
Random noise or spike noise may be due to errors during transmission of
data or to a temporary disturbance. Here, individual pixels acquire DN-values
that are much higher or lower than the surrounding pixels (Figure 8.6). In the
image, these pixels produce bright and dark spots that interfere with information extraction procedures.
A spike noise can be detected by mutually comparing neighbouring pixel
values. If neighbouring pixel values differ by more than a specific threshold
margin, it is designated as a spike noise and the DN is replaced by an interpolated DN-value.
34
28
23
17
13
11
10
10
10
10
15
25
33
39
43
46
48
50
51
52
53

27
21
17
14
12
11
11
11
11
11
18
31
40
44
47
48
50
52
53
54
55

20
16
14
13
12
11
180
11
11
13
21
36
45
47
49
50
52
120
55
56
57

17
15
14
12
11
11
11
11
13
17
26
40
47
49
50
51
53
55
57
58
59

17
17
15
13
11
11
11
13
17
25
33
42
48
50
51
53
55
57
58
58
58

19
18
16
12
11
11
12
16
23
33
40
46
49
51
52
54
0
58
59
59
58

20
18
15
11
10
12
16
22
30
40
47
49
51
53
54
55
56
58
59
59
58

21
18
15
12
12
15
20
28
36
45
50
50
51
53
54
54
55
56
57
57
57

22
20
17
14
16
22
28
36
42
47
50
50
50
51
52
52
53
54
55
55
55

21
204
16
13
15
23
31
39
45
48
50
50
50
50
51
51
51
51
52
52
52

19
16
13
11
13
21
29
39
45
48
50
50
8
50
51
51
50
50
49
49
48

18
14
12
12
16
24
32
42
48
50
51
51
52
52
51
51
50
50
49
49
49

16
14
14
17
23
31
39
47
51
53
54
54
54
54
53
53
52
52
51
51
52

17
17
19
25
31
39
46
52
55
55
55
55
55
55
55
55
55
55
54
2
55

22
24
28
34
40
46
51
55
57
57
57
57
57
57
57
57
57
58
59
59
61

(b)

(a)

first

last

back

exit

zoom

contents

index

about

Figure 8.6: The image

with spike errors (a) and
the DN-values (b).

243

8.3. Atmospheric Corrections

8.3

Atmospheric Corrections

All reflected and emitted radiations leaving the Earths surface are attenuated
mainly due to absorption and scattering by the constituents in the atmosphere
(refer to Section 2.1). The atmospheric induced distortions occur twice in case of
sunlight reflection and once in case of emitted radiation. These distortions are
wavelength dependent. Their effect on remote sensing data can be reduced by
applying atmospheric correction techniques. These corrections are related to
the influence of
haze,
sun angle, and
skylight.

first

last

back

exit

zoom

contents

index

about

244

8.3. Atmospheric Corrections

8.3.1 Haze correction

Light scattered by the atmospheric constituents that reaches the sensor constitutes the haze in remote sensing image data. Haze has an additive effect resulting in higher DN-values and a decrease in the overall contrast in the image
data. The effect is wavelength dependent, being more pronounced in the shorter
wavelength range and negligible in the infrared.
Haze corrections are based on the assumption that the infrared bands are
essentially free of atmospheric effects and in these bands black bodies, such as
large clear water bodies and shadow zones, will have zero DN-value. The DNvalues in other bands for the corresponding pixels can be attributed to haze and
should be subtracted from all pixels of the corresponding band.

first

last

back

exit

zoom

contents

index

about

245

8.3. Atmospheric Corrections

8.3.2 Sun angle correction

The position of the sun, relative to the earth, changes depending on the time of
the day and the day of the year. In the northern hemisphere, the solar elevation
angle is smaller in winter than in summer (see Figure 8.7). As a result, the image
data of different seasons are acquired under different solar illumination. Sun
angle correction becomes more important when one wants to generate mosaics
taken at different times or perform change detection studies.

Satellite

Ta
en

ith

lan

n
Ze

Equator

Sun angle

Sun

Summer

Sun

Spring
Autumn

Sun

Winter

Figure 8.7: Effects of seasonal changes on the solar elevation angle in the
northern hemisphere (after Lillesand and Kiefer,
1994).

An absolute correction involves dividing the DN-values in the image data by

the sine of the solar elevation angle (the size of the angle is given in the header

first

last

back

exit

zoom

contents

index

about

246

8.3. Atmospheric Corrections

of the image data), as per the following formula.
DN 0 =

DN
sin()

(8.1)

Here, DN is the input pixel value, DN 0 is the output pixel value, and is the
solar elevation angle. Note that since the angle is smaller than 90 , the sine will
be smaller than 1 and DN will be larger than DN.
When multitemporal data sets of the same area are available, a relative sun
angle correction can be performed. In such cases, the image with higher sun
elevation angle is taken as a reference and the radiometric values of the other
image are adjusted to it.

The following are typical exam questions:
1. Which are the radiometric errors that can be introduced due to malfunctioning of satellite sensors?
2. What are the differences between line dropouts and line striping?
3. Explain the procedure for haze correction on remotely sensed data.

first

last

back

exit

zoom

contents

index

about

Chapter 9
Geometric aspects

first

last

back

exit

zoom

contents

index

about

252

253

9.1. Introduction

9.1

Introduction

The discussions of the various sensors in chapters 4, 5, and 6 have indicated

some of the geometric characteristics inherent to different type of image data.
These geometric characteristics have to be taken into consideration when the
data are used:
To derive two-dimensional (x, y) and three-dimensional (x, y, z) coordinate
information. 2D geometric descriptions of objects (points, lines, areas) can
be derived from a single image or photo. 3D geometric descriptions (2.5D
terrain relief, 3D objects as volumes) can be derived from stereo pairs of
images or photos. Extraction of 3D information from images requires a
specific process called orientation.
To merge different types of image data for integrated processing and analysis. Consider a land cover classification based on multispectral Landsat
and SPOT data. Both data sets need to be converted into the same geometric grid before they can be processed simultaneously. This can be achieved
by a geocoding process.
To visualize the image data in a GIS environment. There is a growing
amount of image data available that are used as a backdrop for other (vector stored) data. To enable such integration, the image data need to be
georeferenced to the coordinate system of the vector stored data.
Traditionally, the subjects of georeferencing, geocoding and orientation have
been addressed by the discipline of photogrammetry. Today, not only aerial photographs but also various other types of image data are used in the field of digital

first

last

back

exit

zoom

contents

index

about

254

9.1. Introduction
photogrammetry [25]. Deriving 3D measurements from radar data is in the discipline of interferometry and radargrammetry. This chapter deals with a limited
number of concepts and topics. In Section 9.2 relief displacement is introduced.
Section 9.3 and Section 9.4 introduce 2D and 3D approaches respectively. 2D
approaches relate to methods and techniques that neglect terrain elevation differences. In 3D approaches, methods for correction of relief displacement are
introduced. If a DTM is not available, 3D coordinates can be measured directly
from stereo pairs. Note that spatial referencing in general and an introduction
to map projections have been introduced in Principles of GIS [6].

first

last

back

exit

zoom

contents

index

about

255

9.2. Relief displacement

9.2

Relief displacement

A characteristic of most sensor systems is the distortion of the geometric relationship between the image data and the terrain caused by relief differences on
the ground. This effect is most apparent in aerial photographs and airborne
scanner data. The effect of relief displacement is illustrated in Figure 9.1. Consider the situation on the left in which a true vertical aerial photograph is taken
of a flat terrain. The distances (A B) and (a b) are proportional to the total
width of the scene and its image on the negative, respectively. In the left hand
situation, by using the scale factor, we can compute (AB) from a measurement
of (a b) in the negative. In the right hand situation, there is significant terrain
relief difference. As you can now observe, the distance between a and b in the

(a)

first

last

back

exit

zoom

contents

(b)

index

about

Figure 9.1: Illustration of

the effect of terrain topography on the relationship between AB (on the
ground) and ab (on the
photograph). Flat terrain
(a); significant height difference (b).

256

9.2. Relief displacement

negative has become larger, although when measured in the terrain system, it is
still the same as in the left hand situation. This phenomenon does not occur in
the centre of the photo but becomes increasingly prominent towards the edges
of the photo. This effect is called relief displacement: terrain points whose elevation is above or below the reference elevation are displaced respectively away
from or towards the nadir point. The magnitude of displacement, r (mm), is
approximated by:
rh
r =
(9.1)
H
In this equation, r is the radial distance (mm) from the nadir, h (m) is the
height of the terrain above the reference plane and H (m) is the flying height
above the reference plane (where nadir intersects the terrain). The equation
shows that the amount of relief displacement is zero at the nadir (r = 0), greatest at the corners of the photograph and, is inversely proportional to the flying
height.
In addition to relief displacement you can imagine that also buildings and
other tall objects can cause displacement (height displacement). This effect is, for
example, encountered when dealing with large scale photos of urban or forest
areas (Figure 9.2). In this chapter the subject of height displacement is not further
elaborated.
The main effect of relief displacement is that inaccurate or wrong coordinates
might be determined when for example digitizing from image data. Whether relief displacement should be considered in the geometric processing of the image
data depends on its impact on the required accuracy of the geometric information derived from the images. Relief displacement can be corrected for if information on the terrain topography is available (in the form of a DTM). The
procedure is explained in Section 9.4.
first

last

back

exit

zoom

contents

index

about

257

9.2. Relief displacement

Figure 9.2: Fragment of a

large scale aerial photograph of the centre of Enschede. Note the effect
of height displacement on
the higher buildings.

first

last

back

exit

zoom

contents

index

about

258

9.3. Two-dimensional approaches

9.3

Two-dimensional approaches

In this section we consider the geometric processing of image data in situations

where relief displacement can be neglected. Examples of such image data are a
digitized aerial photograph of a flat area (flat being defined by h/H < 1/1000).
For images taken from space with only a medium resolution, the relief displacement usually is less than a few pixels and thus not important. These data are
stored in a column-row system in which columns and rows are indicated by index i and j respectively. The objective is to relate the image coordinate system
to a specific map coordinate system (Figure 9.3).

Column (i)
1
Row
(j)

4
y

(x,y)

2
3
4

(i,j)

(0,0)

(a)
first

last

back

exit

zoom

contents

index

(b)
about

Figure 9.3:
Coordinate
system of the image defined by rows and columns
(a) and map coordinate
system with x- and y-axes
(b).

259

9.3. Two-dimensional approaches

9.3.1 Georeferencing
The simplest way to link an image to a map projection system is to use a geometric
transformation. A transformation is a function that relates the coordinates of two
systems. A transformation, relating (x, y) to (i, j) is typically defined by linear
equations, such as: x = 3 + 5i, and, y = 2 + 2.5j.
Using the above transformation, for example, image position (i = 5, j = 8)
relates to map coordinates (x = 28, y = 18). Once such a transformation has
been determined, the map coordinates for each image pixel can be calculated.
The resulting image is called a georeferenced image. It allows the superimposition
of vector data and the storage of the data in map coordinates when applying
on-screen digitizing. Note that the image as such remains stored in the original
(i, j) raster structure, and that its geometry is not altered.
The process of georeferencing includes two steps: selection of the appropriate type of transformation, and determination of the transformation parameters.
The type of transformation depends mainly on the sensor-platform system used.
For aerial photographs (of a flat terrain) usually a so-called perspective transformation is used to correct for the effect of tilt and roll (Section 3.2). A more general
transformation is the polynomial transformation, which enables 1st , 2nd to nth order transformations. In many situations a 1st order transformation is adequate.
A 1st order transformation relates map coordinates (x, y) with image coordinates
(i, j) as follows:
x = a + bi + cj

(9.2)

y = d + ei + f j

(9.3)

Equations 9.2 and 9.3 require that six parameters (a . . . f ) be determined. The
transformation parameters can be determined by means of ground control points
first

last

back

exit

zoom

contents

index

about

260

9.3. Two-dimensional approaches

GCP
1
2
3
4
5

i
254
149
40
26
193

j
68
22
132
269
228

x
958
936
916
923
954

y
155
151
176
206
189

xc
958.552
934.576
917.732
921.835
954.146

yc
154.935
150.401
177.087
204.966
189.459

dx
0.552
-1.424
1.732
-1.165
0.146

Table 9.1: A set of five

ground control points,
which are used to determine a 1st order
transformation. xc and yc
are calculated using the
transformation, dx and dy
are the residual errors.

dy
-0.065
-0.599
1.087
-1.034
0.459

(GCPs). GCPs are points that can be clearly identified in the image and in a
source, which is in the required map projection system. One possibility is to use
topographical maps of an adequate scale. The operator then needs to identify
identical points on both sources e.g., using road crossings, waterways, typical
morphological structures, et cetera. Another possibility is to identify points in
the image and to measure the coordinates in the field by satellite positioning.
The result of GCP selection is a set of related points (Table 9.1). To solve the
above equations, at least three GCPs are required. However, to calculate the
error of the transformation more points are required. Based on the set of GCPs
the computer calculates the optimal transformation parameters using a best fit
procedure. The errors that remain after transformation are called residual errors.
Their magnitude is an indicator of the quality of the transformation.
Table 9.1 shows an example in which five GCPs have been used. Based on
the given (i, j) and (x, y), one can determine the following transformation:
x = 902.76 + 0.206i + 0.051j,
and
y = 152.579 0.044i + 0.199j.
first

last

back

exit

zoom

contents

index

about

261

9.3. Two-dimensional approaches

Using these equations, you can calculate the image coordinates for each individual GCP. In doing so, you will not exactly find the x and y values in the
table because you calculate with three decimals only. Furthermore, a difference
remains between the calculated values, (xc , yc ), and true values, (x, y), because
the transformation has been optimized to give the best fit for the whole set of
GCPs. The resulting differences (residual errors) are listed by dx and dy . The
residual errors can be used to analyse which GCPs have the largest contribution
to the errors. This may indicate, for example, a GCP that has been inaccurately
identified.
The overall accuracy of a transformation is usually expressed by the Root
Mean Square Error (RMS error), which calculates a mean value from the individual residuals. The RMS error in x-direction, mx , is calculated using the following
equation:
v
u
n
u1 X
xi 2
mx = t

(9.4)

n i=1

For the y-direction, a similar equation can be used to calculate my . The overall
error, mtotal , is calculated by:
q

mtotal =

mx 2 + my 2

(9.5)

For the example data set given in Table 9.1, the residuals have been calculated. The respective mx , my and mtotal are 1.159, 0.752 and 1.381. The RMS error
is a quantitative method to check the accuracy of the transformation. However,
the RMS error does not take into account the spatial distribution of the GCPs.
The RMS error is only valid for the area that is bounded by the GCPs. In the
selection of GCPs, therefore, you should preferably include points located near
the edges of an image.
first

last

back

exit

zoom

contents

index

about

262

9.4

Three-dimensional approaches

Unlike the previous section, which was concerned only with 2D coordinates, in
this section also the 3rd dimension is relevant. The following processes will be
explained in the subsequent sections:
Monoplotting, which is an approach to correct for terrain relief during digitizing of terrain features from aerial photos that results in relatively accurate (x, y) coordinates.
Orthoimage production, which is an approach to correct image data for terrain relief and store the image in a specific map projection. Orthoproducts
can be used as a backdrop to other data or, used to directly determine the
(x, y) geometry of the features of interest. Today, many organizations and
companies sell orthoproducts.
Stereoplotting, which is used to extract 3D information from stereo pairs.
Stereo pairs allow for the determination of (x, y, z) positions. 3D information extraction requires an orientation process. Examples of 3D products
are a Digital Terrain Model (DTM) and large-scale databases, for instance,
related to urban constructions, roads, or cadastral parcels.

first

last

back

exit

zoom

contents

index

about

266

9.4. Three-dimensional approaches

Aerial photo

Digital terrain model

Ground control
points

(x,y)

Figure 9.6: The process

of digital monoplotting enables accurate determination of terrain coordinates
from a single aerial photograph.

Processing

(X,Y,Z)

9.4.1 Monoplotting
Suppose you need to derive accurate positions from an aerial photograph expressed in a specific map projection. This can be achieved for a flat terrain using
a true vertical photograph and a georeferencing approach. However, if there are
significant terrain relief differences you need to correct for relief displacement.
For this purpose the method of monoplotting has been developed.
Monoplotting is based on reconstruction of the position of the camera at the
moment of exposure relative to a Digital Terrain Model (DTM) of the terrain.
This is achieved by identification of a number (at least four) of ground control
points (GCPs) for which both the photo and map coordinates must be known.
The applied DTM should be stored in the required map projection system and
the heights should be expressed in an adequate vertical reference system. When
digitizing features from the photograph, the computer uses the DTM to calculate

first

last

back

exit

zoom

contents

index

about

267

9.4. Three-dimensional approaches

the effect of relief displacement and corrects for it. A monoplotting approach
is possible by using a hard-copy image on a digitizer tablet, or by on-screen
digitizing from a digital image. In the latter situation, vector-stored information
can be superimposed over the image to update the changed features. Note that
monoplotting is a (real time) correction procedure and does not yield new image
data.

Distance between tree tops
Distance between tree bases

Figure 9.7: The same tree

about

Chapter 10
Image enhancement and
visualisation

first

last

back

exit

zoom

contents

index

about

275

276

10.1. Introduction

10.1

Introduction

Many of the figures in the previous chapters have presented examples of remote
sensing image data. There is a need to visualize image data at most stages of
the remote sensing process. For example, the procedures for georeferencing, explained in Chapter 8, cannot be performed without visual examination to measure the location of ground control points on the image. However, it is in the
process of information extraction that visualization plays the most important
role. This is particularly so in the case of visual interpretation (Chapter 11), but
also during automated classification procedures (Chapter 12).
Because many remote sensing projects make use of multispectral data, this
chapter focuses on the visualization of colour imagery. An understanding of
how we perceive colour is required at two main stages in the remote sensing
process. In the first instance, it is required in order to produce optimal pictures from (multispectral) image data on the computer screen or as a (printed)
hard-copy. Thereafter, the theory of colour perception plays an important role in
the subsequent interpretation of these pictures. To understand how we perceive
colour, Section 10.2 deals with the theory of colour perception and colour definition. Section 10.3 gives the basic principles you need to understand and interpret
the colours of a displayed image. The last section (Section 10.5) introduces some
filter operations for enhancing specific characteristics of the image.

first

last

back

exit

zoom

contents

index

about

277

10.2. Perception of colour

10.2

Perception of colour

Colour perception takes place in the human eye and the associated part of the
brain. Colour perception concerns our ability to identify and distinguish colours,
which, in turn, enables us to identify and distinguish entities in the real world.
It is not completely known how human vision works, or what exactly happens
in the eyes and brain before someone decides that an object is (for example) light
blue. Some theoretical models, supported by experimental results are, however,
generally accepted. Colour perception theory is applied whenever colours are
reproduced, for example in colour photography, TV, printing and computer animation.

first

last

back

exit

zoom

contents

index

about

278

10.2. Perception of colour

10.2.1 Tri-stimuli model

The eyes general sensitivity is to wavelengths between 400700 nm. Different
wavelengths in this range are experienced as different colours.
The retinas in our eyes have cones (light-sensitive receptors) that send signals
to the brain when they are hit by photons with energy levels that correspond
to different wavelengths in the visible range of the electromagnetic spectrum.
There are three different kinds of cones, responding to blue, green and red wavelengths (Figure 10.1). The signals sent to our brain by these cones, and the differences between them, give us colour-sensations. In addition to cones, we have
rods, which do not contribute to colour vision. The rods can operate with less
light than the cones. For this reason, objects appear less colourful in low light
conditions.
eye sensitivity

Blue

400

wavelength (nm)

Green Red

500

600

Figure 10.1: Visible range

of electromagnetic spectrum including the sensitivity curves of cones in the
human eye.

700

10.2. Perception of colour

RGB
The RGB definition of colours is directly related to the way in which computer
and television monitors function. Three channels (RGB) directly related to the
red, green and blue dots are input to the monitor. When we look at the result,
our brain combines the stimuli from the red, green and blue dots and enables
us to perceive all possible colours from the visible part of the spectrum. During
the combination, the three colours are added. When green dots are illuminated
in addition to red ones, we see yellow. This principle is called the additive colour
scheme. Figure 10.2 illustrates the additive colours caused by bundles of light
from red, green and blue spotlights shining on a white wall in a dark room.
When only red and green light occurs, the result is yellow. In the central area
there are equal amounts of light from all three the spotlights, and we experience
white.

(a)

Figure 10.2: Comparison

of the (a) additive and
(b) subtractive colour
schemes

(b)

In the additive colour scheme, all visible colours can be expressed as combinations of red, green and blue, and can therefore be plotted in a three-dimensional
space with R, G and B along the axes. The space is bounded by minimum and
maximum values (intensities) for red, green and blue, defining the so-called
first

last

back

exit

zoom

contents

index

about

282

10.2. Perception of colour

[0,1,0] green

yellow
[1,1,0]
white

cyan
[0,1,1]

ic
at

lin

[1,1,1]

[0,0,0]

[0,0,1]

o
hr
ac

red

black

[1,0,0]

Figure 10.3: The RGB

cube; note the red, green
and blue corner points

magenta
[1,0,1]

blue

colour cube (Figure 10.3).

first

last

10.3

Visualization of image data

In this section, various ways for visualizing single and multi-band image data
are introduced. The section starts with an explanation of the concept of an image histogram. The histogram has a crucial role in realizing optimal contrast of
images. An advanced section deals with the application of RGB-IHS transformation to integrate different types of image data.

first

last

back

exit

zoom

contents

index

about

287

10.3. Visualization of image data

10.3.1 Histograms
A number of important characteristics of a single-band image, such as a panchromatic satellite image, a scanned monochrome photograph or a single band from
a multi-band image, are found in the histogram of that image. The histogram describes the distribution of the pixel values (Digital Numbers, DN) of that image.
In the usual case, the DN-values range between 0255. A histogram indicates
the number of pixels for each value in this range. In other words, the histogram
contains the frequencies of DN-values in an image. Histogram data can be represented either in tabular form or graphically. The tabular representation (Table 10.1)
normally shows five columns. From left to right these are:
DN: Digital Numbers, in the range [0. . . 255]
Npix: the number of pixels in the image with this DN (frequency)
Perc: frequency as a percentage of the total number of image pixels
CumNpix: cumulative number of pixels in the image with values less than
or equal to DN
CumPerc: cumulative frequency as a percentage of the total number of
image pixels
Histogram data can be further summarized in some characteristic statistics:
mean, standard deviation, minimum and maximum, as well as the 1% and 99%
values (Table 10.2). Standard deviation is a statistical measure of the spread of
the values around the mean. The 1% value, for example, defines the cut-off value
below which only 1% of all the values are found. 1% and 99% values can be used
to define an optimal stretch for visualization.
first

last

back

exit

zoom

contents

index

about

288

10.3. Visualization of image data

first

DN
0
13
14
15
16
51
52
53
54
102
103
104
105
106
107
108
109
110
111
163
164
165
166
173
174
255

Npix
0
0
1
3
2
55
59
94
138
1392
1719
1162
1332
1491
1685
1399
1199
1488
1460
720
597
416
274
3
0
0

last

Perc
0.00
0.00
0.00
0.00
0.00
0.08
0.08
0.13
0.19
1.90
2.35
1.59
1.82
2.04
2.31
1.91
1.64
2.04
2.00
0.98
0.82
0.57
0.37
0.00
0.00
0.00

back

CumNpix
0
0
1
4
6
627
686
780
918
25118
26837
27999
29331
30822
32507
33906
35105
36593
38053
71461
72058
72474
72748
73100
73100
73100

exit

Table 10.1: Example histogram in tabular format

CumPerc
0.00
0.00
0.00
0.01
0.01
0.86
0.94
1.07
1.26
34.36
36.71
38.30
40.12
42.16
44.47
46.38
48.02
50.06
52.06
97.76
98.57
99.14
99.52
100.00
100.00
100.00

zoom

contents

index

about

289

10.3. Visualization of image data

Mean
113.79

StdDev
27.84

Min
14

Max
173

1%-value
53

Table 10.2:
Summary
statistics for the example
histogram given above.

99%-value
165

In addition to the frequency distribution, the graphical representation shows

the cumulative frequency (see Figure 10.5). The cumulative frequency curve
shows the percentage of pixels with DN that are less than or equal to a given
value.
100%

Cumulative histogram

2%
Ordinary histogram

50%

first

0 14

114

last

back

0%
255

173

exit

Figure 10.5:
Standard
histogram and cumulative
histogram corresponding
with Table 10.1.

zoom

contents

index

about

290

10.3. Visualization of image data

10.3.2 Single band image display

The histogram is used to obtain optimum display of single band images. Single band images are normally displayed using a grey scale. Grey shades of the
monitor typically range from black (value 0) to white (value 255). When applying grey shades, the same signal is input to each of the three (RGB) channels of
the computer monitor (Figure 10.6).

Figure 10.6: Multi-band

image displayed on a monitor using the monitors
Red, Green and Blue input
channels.

Using the original image values to control the monitor values usually results in an image with little contrast since only a limited number of grey values
are used. In the example introduced in the previous section (Table 10.2) only
173 14 = 159 out of 255 grey levels would be used. To optimize the range of
grey values, a transfer function maps DN-values into grey shades on the monitor
(Figure 10.7). The transfer function can be chosen in a number of ways. Linear
contrast stretch is obtained by finding the DN-values where the cumulative histogram of the image passes 1% and 99%. DNs below the 1% value become black
(0), DNs above the 99% value are white (255), and grey levels for the intermediate values are found by linear interpolation. Histogram equalization, or histogram
stretch, shapes the transfer function according to the cumulative histogram. As
a result, the DNs in the image are distributed as equally as possible over the
available grey levels (Figure 10.7).
first

last

back

exit

zoom

contents

index

about

291

10.3. Visualization of image data

255
(white)

no stretch

linear stretch

histogram stretch

Figure 10.7: The relationship between image value

and monitor grey value
is defined by the transfer
function, which can have
different shapes.

grey
shade
(black)
0

255

Transfer function manipulations are usually implemented as modifications

band 1 (blue)
2 (green)
3 (red)
4 (near-ir)
5 (near-ir)
6 (thermal-ir)
7 (mid-ir)

Natural colour composite

(3,2,1)

False colour composite

(4,3,2)

Pseudo-natural colour
composite (3,5,2)

first

last

back

Figure 10.9: Landsat TM

false colour composite
of Enschede and surrounding. Three different
colour composites are
shown:
natural colour,
pseudo-natural colour and
false colour composite.
Other band combinations
are possible.

exit

zoom

contents

index

about

295

10.4. Colour composites

10.4.1 Application of RGB and IHS for image fusion

One of the practical applications of (switching) between different colour spaces
is in the fusion of image data with different spectral and spatial resolutions. A
colour image with three bands for red, green and blue can actually be transformed into the IHS domain and back again. This yields a new image, again
with three components. When looking at these components separately, the intensity image gives a proper grey scale representation of the colour image. The
hue image and the saturation image are not very informative, although they can
be manipulated. For example, when all saturation values appear low, it is possible to increase the saturation by a certain amount. When transforming the hue,
intensity and the new saturation image back to the RGB domain, saturation enhancement is achieved. An example application of this method in image fusion is
described below.
Consider the following two images:
SPOT XS multispectral image with 20 m pixels.
SPOT panchromatic image with 10 m pixels.
In certain applications, we would like to have colour information at a high (10 m)
resolution. It is, however, not available from SPOT XS. The closest we can get is
to create an image that combines hue and saturation from XS with intensity from
Pan (Figure 10.10). First step is to resample the SPOT XS (pixels of 20 m) into the
image raster of the SPOT Pan (pixels of 10 m). The SPOT XS is then converted
from (RGB) to (IHS). Subsequently, I is replaced by the Pan image, and (Pan,
H, S) converted back to (RGB). The result is a high-resolution (10 m) colour
composite, which is potentially more suitable for image interpretation purposes
than the two separate originals. The same technique can be used to fuse other
first

last

back

exit

zoom

contents

index

about

CAUTION

296

10.4. Colour composites

SPOT XS
band 3

SPOT XS
band 2

SPOT XS
band 1

SPOT
Pan

Resampling and (optional) contrast stretch

combinations of data, for example TM with SPOT PAN, SPOT XS with ERS-SAR,
multispectral satellite imagery with B/W aerial photography, et cetera.
Red

Green

Hue

RGB to
HSI

Blue

Red

HSI
to RGB

Satur.

Int.

Green

Color
composite

Blue

Figure 10.10: Procedure

to merge SPOT panchromatic and multispectral
data using RGB to IHS
conversion and vice versa.

contrast
stretch

Intensity

Fusion of SPOT XS and SPOT Panchromatic into 10 m resolution color composite.

first

last

back

exit

zoom

contents

index

about

297

10.5. Filter operations

10.5

Filter operations

A further step in producing optimal images for interpretation is the use of filter
operations. Filter operations are local image transformations: a new image is
calculated and the value of a pixel depends on the values of its former neighbours. Filter operations are usually carried out on a single band. Filters are
used for spatial image enhancement, for example to reduce noise or to sharpen
blurred images. Filter operations are extensively used in various semi-automatic
procedures that are outside the scope of this chapter.
To define a filter, a kernel is used. A kernel defines the output pixel value as a
linear combination of pixel values in a neighbourhood around the corresponding position in the input image. For a specific kernel, a so-called gain can be
calculated as follows:
gain =

1
ki

(10.1)

The gain sums all kernel coefficients (ki ). In general, the sum of the kernel
coefficient, after multiplication by the gain, should be equal to 1 to result in an
Figure 10.11: Input and
output result of a filtering
operation: the neighbourhood in the original image determines the value
of the output. In this situation a smoothing filter was
applied.

Output

Input

16 12 20

first

last

back

exit

zoom

contents

index

about

298

10.5. Filter operations

image with approximately the same range of grey values. The effect of using a
kernel is illustrated in Figure 10.11, which shows how the output value is calculated in terms of average filtering.
The significance of the gain factor is explained in the next two subsections.
In these examples only small neighbourhoods of 3 3 kernels are considered. In
practice other kernel dimensions may be used.

first

last

back

exit

zoom

contents

index

about

299

10.5. Filter operations

10.5.1 Noise reduction

Consider the kernel shown in Table 10.3 in which all coefficients equal 1. This
means that the values of the nine pixels in the neighbourhood are summed. Subsequently, the result is divided by 9 to achieve that the overall pixel values in the
output image are in the same range as the input image. In this situation the gain
is 1/9 = 0.11. The effect of applying this averaging filter is that image will become
blurred or smoothed. When dealing with speckle effect in radar imagery the
result of applying this filter is to reduce the speckle.
1
1
1

1
1
1

Table 10.3: Filter kernel

for smoothing.

1
1
1

In the above kernel, all pixels have equal contribution in the calculation of the
result. It is also possible to define a weighted average. To emphasize the value of
the central pixel, a larger value can be put in the centre of the kernel. As a result,
less drastic blurring takes place. In addition, it is necessary to take into account
that the horizontal and vertical neighbours influence the result more strongly
that the diagonal ones. The reason for this is that the direct neighbours are closer
to the central pixel. The resulting kernel, for which the gain is 1/16 = 0.0625, is
given in Table 10.4.
1
2
1

first

last

2
4
2

back

Table 10.4: Filter kernel

for weighted smoothing.

1
2
1

exit

zoom

contents

index

about

300

10.5. Filter operations

10.5.2 Edge enhancement

Another application of filtering is to emphasize local differences in grey values,
for example related to linear features such as roads, canals, geological faults, et
cetera. This is done using an edge enhancing filter, which calculates the difference between the central pixel and its neighbours. This is implemented using
negative values for the non-central kernel coefficients. An example of an edge
enhancement filter is given in Table 10.5.
-1
-1
-1

-1
16
-1

Table 10.5: Filter kernel

used for edge enhancement.

-1
-1
-1

The gain is calculated as follows: 1/(16 8) = 1/8 = 0.125. The sharpening

effect can be made stronger by using smaller values for the centre pixel (with a
minimum of 9). An example of the effect of using smoothing and edge enhancement is shown in Figure 10.12.

Figure 10.12: Original image (middle), edge enhanced image (left) and
smoothed image (right).

first

last

back

exit

zoom

contents

index

about

301

10.5. Filter operations

Summary
The way we perceive colour is most intuitively described by the Hue component
of the IHS colour space. The colour space used to describe colours on computers
monitor is the RGB space.
When displaying an image on a screen (or on hard copy) many choices need
to be made. The selection of bands, the sequence in which these are linked to the
Red-Green-Blue channels of the monitor, the use of stretching techniques and
the possible use of (spatial) filtering techniques.
The histogram, and the derived cumulative histogram, is the basis for all
stretching methods. Stretching, or contrast enhancement, is realized using transfer functions.
Filter operations are based on the use of a kernel. The weights of the coefficients in the kernel determine the effect of the filter which can be, for example,
to smooth or sharpen the original image.

first

last

back

exit

zoom

contents

index

about

302

10.5. Filter operations

Questions
The following questions can help you to study Chapter 10.
1. How many possibilities are there to visualize a 4 band image using a computer monitor?
2. You are shown a picture in which grass looks green and houses are red
what is your conclusion? Now, you are shown a picture in which grass
shows as purple and houses are blackwhat is your conclusion now?
3. What would be a reason for not using the default application of histogram
equalization for all image data?

first

last

back

exit

zoom

contents

index

about

303

10.5. Filter operations

4. Can you think of a situation in your own context where you would probably use filters to optimize interpretation of image data?

first

last

back

exit

zoom

contents

index

about

304

10.5. Filter operations

The following are typical exam questions:
1. List the three colour spaces used in the context of remote sensing and visualization.
2. Which colour space should be applied when using computer monitors?
How is the colour white produced?
3. What information is contained in a histogram of image data?

first

last

back

exit

zoom

contents

index

about

305

10.5. Filter operations

4. Which technique is used to maximize the range of colours (or grey values)
when displaying an image?
5. Using an example, explain how a filter works.

first

last

back

exit

zoom

contents

index

about

Chapter 11
Visual image interpretation

first

last

back

exit

zoom

contents

index

about

306

307

11.1. Introduction

11.1

Introduction

Up to now, we have been dealing with acquisition of image data. The data acquired still needs to be interpreted (or analysed) to extract the required information. In general, information extraction methods from remote sensing imagery
can be subdivided into two groups:
Information extraction based on visual analysis or interpretation of the
data. Typical examples of this approach are visual interpretation methods for land use or soil mapping. Also the generation/updating of topographic maps from aerial photographs is based on visual interpretation.
Visual image interpretation is introduced in this Chapter.
Information extraction based on semi-automatic processing by the computer. Examples include automatic generation of DTMs, image classification and calculation of surface parameters. Image classification is introduced
in Chapter 12.
The most intuitive way to extract information from remote sensing images
is by visual image interpretation, which is based on mans ability to relate colours
and patterns in an image to real world features. Chapter 10 has explained different methods used to visualize remote sensing image data.
In some situations pictures are studied to find evidence of the presence of
features, for example, to study natural vegetation patterns. Most often the result
of the interpretation is made explicit by digitizing the geometric and thematic
data of relevant objects (mapping). The digitizing of 2D features (points, lines
and areas) is carried out using a digitizer tablet or on-screen digitizing. 3D features interpreted in stereopairs can be digitized using stereoplotters or digital
photogrammetric workstations.
first

last

back

exit

zoom

contents

index

about

308

11.1. Introduction
In Section 11.2 some theory about image understanding is explained. Visual
image interpretation is used to produce spatial information in all of ITCs fields
of interest: urban mapping, soil mapping, geomorphological mapping, forest
mapping, natural vegetation mapping, cadastral mapping, land use mapping
and many others. As visual image interpretation is application specific, it is illustrated by two examples (soil mapping, land cover mapping) in (Section 11.3).
The last section (11.4) addresses some aspects of quality.

first

last

back

exit

zoom

contents

index

about

309

11.2. Image understanding and interpretation

11.2

first

Image understanding and interpretation

last

back

exit

zoom

contents

index

about

310

11.2. Image understanding and interpretation

11.2.1 Human vision

In Chapter 10 humans perception of colour was explained. Human vision goes
a step beyond perception of colour: it deals with the ability of a person to draw
conclusions from visual observations. In analysing a picture, typically you are
somewhere between the following two situations: direct and spontaneous recognition or using several clues to draw conclusions by a reasoning process (logical
inference).

Figure 11.1: Satellite image of Antequera area in

Spain, the circular features
are pivot irrigation systems

Spontaneous recognition refers to the ability of an interpreter to identify objects

or phenomena at a first glance. Consider Figure 11.1. An agronomist would immediately recognize the pivot irrigation systems with their circular shape. S/he
would be able to do so because of earlier (professional) experience. Similarly,
most people can directly relate an aerial photo to their local environment. The

first

last

back

exit

zoom

contents

index

about

311

11.2. Image understanding and interpretation

quote from people that are shown an aerial photograph for the first time I see
because I know refers to spontaneous recognition.
Logical inference means that the interpreter applies reasoning. In the reasoning
the interpreter will use his/her professional knowledge and experience. Logical
inference is, for example, concluding that a rectangular shape is swimming pool
because of its location in a garden and near to a house. Sometimes, logical inference alone cannot help you in interpreting images and field observations are
required (Section 11.4). Consider the aerial photograph in Figure 11.2. Would
you be able to interpret the material and function of the white mushroom like
objects? A field visit would be required for most of us to relate the different
features to elements of a house or settlement.

Figure 11.2: Mud huts

of Labbezanga near the
Niger river.
Photo by
Georg Gerster, 1972.

first

recognition of high chimneys, large buildings, cooling towers, coal heaps
and transportation belts. In the same way the land use pattern associated
with small scale farming will be characteristically different to that of large
scale farming.
Having introduced these seven interpretation elements you may have noticed a relation with the spatial extent of the feature to which they relate. Tone
or hue can be defined for a single pixel; texture is defined for a neighbouring
group of pixels, not for single pixels. The other interpretation elements relate to
individual objects or a combination of objects. The simultaneous and often implicit use of all these elements is the strength of visual image interpretation. In
standard image classification (Chapter 12) only hue is applied, which explains
the limitations of automated methods compared to visual image interpretation.

first

last

back

exit

zoom

contents

index

about

315

11.2. Image understanding and interpretation

11.2.3 Stereoscopic vision

The impression of depth encountered in the real world can also be realized by
images of the same object that are taken from different positions. Such a pair
of images, photographs or digital images that are separated and observed at the
same time by the two eyes. These give images on the retinas in which objects
at different positions in space are projected on relative different positions. We
call this stereoscopic vision. Pairs of images that can be viewed stereoscopically
are called stereograms . Stereoscopic vision is explained here because the impression of height and height differences is important in the interpretation of both
natural and man-made features from image data. Note that in Chapter 9 we explained that under specific conditions stereo-models can be used to derive 3D
coordinates.
Under normal conditions we can focuss on objects between 150 mm distance
and infinity. In doing so we direct both eyes to the object (point) of interest. This
is known as convergence. To view the stereoscopic model formed by a pair of
overlapping photographs, the two images have to be separated so that the left
and right eyes see the left and right photographs. In addition one should not focus on the photo itself but at infinity. Some experienced persons can experience
stereo by putting the two photos at a suitable distance from their eyes. Most of
us need some help and different methods have been developed.
Pocket and mirror stereoscopes, and also the photogrammetric plotters, use a
system of lenses and mirrors to feed one image into one eye. Pocket and mirror
stereoscopes are mainly applied in mapping applications related to vegetation,
forest, soil and geomorphology (Figure 11.3) Photogrammetric plotters are used
in topographic and large scale mapping activities.
Another way of achieving stereovision is to project the two images in two
colours. Most often red and green colours are applied; the corresponding specfirst

last

back

exit

zoom

contents

index

about

316

11.2. Image understanding and interpretation

Figure 11.3: The mirror

stereoscope
enables
stereoscopic vision of
stereograms. Each photo
is projected onto one eye.

tacles comprise one red and one green glass. This method is known as the
anaglyph system and is particularly suited to viewing overlapping images on
a computer screen. An approach used in digital photogrammetric systems is to
apply polarization for the left and right images. Polarized spectacles make the
left image visible to the left eye and the right image to the right eye.

first

last

back

exit

zoom

contents

index

about

317

11.3. Application of visual image interpretation

11.3

first

Application of visual image interpretation

back

exit

zoom

contents

index

about

322

11.3. Application of visual image interpretation

the actual class. This provides a quantitative measure of the map accuracy.
Soil mapping is an example of geo-information production with the aid of
aerospace survey, which is based on complementary roles for remote sensing
and ground observations, as well as prior knowledge of the survey area (geology, geomorphology, climate history, ecology, . . . ) and a sound conceptual
framework about the object of study (here, soil forming factors). Remote sensing
is used to provide a synoptic view, based on a mental model of soil-landscape
relationships, and to limit the amount of work in the field by stratification.

first

last

back

exit

zoom

contents

index

about

323

11.3. Application of visual image interpretation

Landscape

Relief

Lithology

Landform

API Code

Hilland

Dissected
ridge

Loess

Summit

Hi111

Shoulder &
backslope
Scarp

Hi112

Toe slope

Hi212

Slope

Hi311

Bottom
Slope

Hi312
Hi411

Tread

Pl311

Escarpment Loess over

basalt
Colluvium
from loess /
basalt
Vales
Alluvium
from loess
Glacis
Plain

High
race

ter-

Old floodplain

first

last

Colluvium
from loess
Loess over
old river alluvium

Old
vium

allu-

back

exit

Table 11.1: Geopedologic

legend, which results in
(hierarchical) API-codes.

Hi211

Abandoned
Pl312
channel
Abandoned
Pl411
floodplain
(channelized)

zoom

contents

index

about

324

11.3. Application of visual image interpretation

API Code

Dominant Soil Type

Hi111
Hi112

Eutri-chromic Cambisols, fine silty

Eutri-chromic Cambisols, coarse silty
Calcaric Regosols, coarse silty
Calcaric Regosols, coarse silty
Stagnic Cambisols, silty skeletal
Eutri-chromic Cambisols, fine silty
Eutric Gleysols, fine silty
Eutric Luvisols, coarse silty
Silti-calcic Kastanozems, fine silty
Mollic Gleysols, fine loamy
Gleyic Fluvisols, fine loamy
Eutric Fluvisols, fine loamy

Hi211
Hi212
Hi311
Hi312
Hi411
Pl311
Pl312
Pl411

first

last

back

exit

Table 11.2: Relationship

between the soil types and
the geo-pedologic legend
is made through the APIcodes.

zoom

60%
40%

50%
50%

contents

index

about

325

11.3. Application of visual image interpretation

11.3.2 Land cover mapping from multispectral data

This second example refers to the CORINE land cover project in which a land
cover database for Europe is being established. CORINE is based on the work
done in different countries of the European Union. This example is to illustrate
a situation in which different organizations and individuals are involved in the
production of the total data set. To enable such approach, research activities
have been carried out that resulted in a technical guide for producing the land
cover data from spaceborne multispectral data [5, 20].
At this point, the terms land cover and land use should be defined since they
both are often used in the context of image interpretation. Land cover refers to
the type of feature present on the surface of the land. It refers to a physical
property or material, e.g., water, sand, potato crop, and asphalt. Land use relates to the human activity or economic function for a specific piece of land, e.g.,
urban use, industrial use or nature reserve. Another way to put it is that land
cover is more dynamic than land use. Most vegetated areas (land cover) change
over time. Land use as it is related to the human activity is more stable: land
is being used for irrigated crops or extensive cattle breeding will not change
within a year. The difference between land cover and land use is also explained
in Figure 11.6 showing that different land use classes can be composed of the
same land cover classes. Principally, remote sensing image data give information about land cover. Using context information the land use sometimes can be
deduced. The distinction between land cover and land use becomes highly relevant in dealing with, for example, environmental impact assessment studies. In
such studies, not only the presence of a certain type of crop is relevant but also
the way in which this crop is grown and treated (in terms of manure, herbicides,
pesticides etc) is important.
The overall CORINE mapping process is as follows:
first

last

back

exit

zoom

contents

index

about

326

11.3. Application of visual image interpretation

The process starts with the selection of cloud free multispectral satellite
images of an area. Hard copy prints are produced from the image data.
Guidelines are given to yield prints that are similar in terms of scale and
colour enhancement. The prints are the basis for the interpretation. (Today,
also on-screen digitizing procedures are described.) Additional information in the form of topographical maps, photo-atlases, soil maps et cetera is
collected to aid the interpretation.
The interpretation of the land cover units is made on transparencies, which
are overlaid on the hard copy images and which are digitized at a later
stage. In the CORINE project the minimum mapping unit is set at 25 ha;
the minimum width of mapping units is set at 100 m. The interpretation is based on the nomenclature, which gives the names of the land
cover classes. CORINE has a three level categorization (Table 11.3 and
Table 11.4). The reason for defining the land cover at three different levels is ensure consistent aggregation of the different classes. At the same
Road
Grass

Buildings

first

Road

Grass

Trees

Figure 11.6: Residential

land use (a) and industrial
land use (b) may be composed of similar land cover
classes.

P
(a)

Trees

last

back

exit

(b)

zoom

contents

index

about

327

11.3. Application of visual image interpretation

time, one can imagine that a map of level-1 classes is more accurate than a
map of level-3 classes. For each class, a more extended class description is
given, followed by instructions on how to recognize the specific class (Table 11.5) together with some example images of the class. The instructions
with respect to discriminating different classes or categories are called an
interpretation key.
The line patterns on the transparencies are digitized and the corresponding
class codes are entered into the computer. The result is a polygon database.
One of the attributes is the land cover code. A sample result of the interpretation is shown in Figure 11.7.
The final activity of any mapping project is an independent assessment of
accuracy (validation). In the CORINE project this is done by field check of
a limited number of objects that are selected by a sampling strategy.

311

112l

231

243

112 l

211

231

l121

231

142
312

242

231

112

211

111l

312

313

Figure 11.7: Example of

the result of the CORINE
Land Cover classification
showing a part around
one of the rivers in The
Netherlands. Courtesy of
Wageningen-UR

231

111

121

l
l

211 l
l

512

111

first

last

111

211
l

l
l121

111

231l

211

242

back

211

exit

zoom

111l

contents

index

about

328

11.3. Application of visual image interpretation

Level 1

1. Artificial

Level 2

1.1.

1.1.1. Continuous urban fab-

Urban fabric

1.1.2.

Surfaces

1.2.

Table 11.3: Part of the

CORINE nomenclature for
land cover.

Level 3

Industrial,
commercial
and transport
units

1.2.1.

ric
Discontinuous
urban
fabric
Industrial or commercial units

1.2.2. Road and rail networks

1.3.

1.4.

Mine,
dump
and construction sites

Artificial nonagricultural
vegetated
areas

1.2.3.
1.2.4.
1.3.1.

and associated land

Port areas
Airports
Mineral extraction sites

1.3.2. Dump sites

1.3.3. Construction sites
1.4.1 Green urban areas

1.4.2. Sport and leisure facilities

first

last

back

exit

zoom

contents

index

about

329

11.3. Application of visual image interpretation

Level 1

Level 2

2. Agricultural 2.1.

Arable land

Table 11.4: Part (two) of

the CORINE nomenclature for land cover.

Level 3

2.1.1. Non-irrigated

areas

arable

land

2.1.2. Permanently

irrigated

land

2.2.

Permanent
crops

2.1.3. Rice fields

2.2.1. Vineyards
2.2.2. Fruit trees and berry
2.2.3.

2.3.
2.4.

Pastures
2.3.1.
Heterogeneous 2.4.1.
agricultural
areas

2.4.2.
2.4.3.

3. Forests

4.
5.
first

and semi
natural
areas
Wetlands
Water bodies
previous

...

2.4.4.
...

...
...

last

back

exit

plantations
Olive groves
Pastures
Annual crops associated with permanent
crops
Complex
cultivation
patterns
Land principally occupied by agriculture, with
significant areas of natural vegetation
Agro-forestry
...

...
...

zoom

contents

index

about

330

11.3. Application of visual image interpretation

Table 11.5:
CORINEs
extended description for
class 1.3.1 (Mineral Extraction Sites). Source:[5]

Class 1.3.1 Mineral Extraction Sites

Extended Description: Areas with open-pit extraction of construction material (sandpits, quarries) or other minerals (open-cast mines). Includes
flooded gravel pits, except for -bed extraction.
How to recognize: Quarries are easily recognizable on satellite images
(white patches) because they contrast with their surroundings. The same
is true for working gravel pits. For open-cast mines, the difference with
item 1.3.2 (dump sites) is not always obvious. In such cases, ancillary
data will be needed to remove any doubt.
Disused open-cast mines, quarries, sandpits, sludge quarries and gravel
pits (not filled with water) are included in this category, However, ruins do
not come under this heading.
Sites being worked or only recently abandoned, with no trace of vegetation, come under this heading. Where vegetal colonization is visible, sites
are classified under the appropriate vegetal cover category.
This heading includes buildings and associated industrial infrastructure
(e.g., cement factories) and small water bodies of less than 25 ha created
by mining.

11.4

Quality aspects

The quality of the result of an image interpretation depends a number of factors:

the interpreter, the image data used and the guidelines provided.
The professional experience and the experience with image interpretation
determine the skills of a photo-interpreter. A professional background is
required: a geological interpretation can only be made by a geologist since
s/he is able to related image features to geological phenomena. Local
knowledge, derived by field visits, is required to help the interpretation.
The image data applied limit the phenomena that can be studied, both
in thematic and geometric sense. One cannot, for example, generate a
reliable database on the tertiary road system using multispectral satellite
data. Likewise, blackandwhite aerial photos contain limited information
about agricultural crops.
Finally, the quality of the interpretation guidelines is of large influence.
Consider, for example, a project in which a group of persons is to carry out
a mapping project. Ambiguous guidelines will prevent a consistent mapping in which individual results form a seamless database of consistent
quality.
Especially in large projects and monitoring programmes, all the above three
points play an important role in ensuring the replicability of the work. Replicability refers to the degree of correspondence obtained by different persons for
the same area or by the same person for the same area at different moments in
time. In principle, replicability does not tell directly about the accuracy (the relation with the real world) but it does tell something about the quality of the class
first

last

back

exit

zoom

contents

index

about

334

11.4. Quality aspects

Figure 11.8: Two interpretation results derived

by two photo-interpreters
analysing the same image. Note the overall differences but also differences in the generalization
of the lines. (from [8])

definition (crisp or ambiguous) and the instructions and methods used. Two
examples are given here to give you an intuitive idea. Figure 11.8 gives two interpretation results for the same area. Note that both results differ in terms of
total number of objects (map units) and in terms of (line) generalization. Figure 11.9 compares 13 individual interpretation results of a geomorphological interpretation. Similar to the previous example, large differences are found along
the boundaries. In addition to this, you also can conclude that for some objects
(map units) there was no agreement on the thematic attribute.

4. In what situations is visual image interpretation preferred to semi-automated interpretations?

5. Give two examples of cases where visual interpretation is not possible,
even with good quality imagery.

first

last

back

exit

zoom

contents

index

about

Chapter 12
Digital image classification

first

last

back

exit

zoom

contents

index

about

341

342

12.1. Introduction

12.1

Introduction

Chapter 11 explained the process of visual image interpretation. In this process,

human vision plays a crucial role in extracting information from image data.
Although computers may be used for visualization and digitization, the interpretation itself is carried out by the operator.
In this chapter, digital image classification is introduced. In this process the
(human) operator instructs the computer to perform an interpretation according to certain conditions. These conditions are defined by the operator. Image
classification is one of the techniques in the domain of digital image interpretation. Other techniques include automatic object recognition (for example, road
detection) and scene reconstruction (for example, generation of 3D object models). Image classification, however, is the most commonly applied technique in
the ITC context.
Application of image classification is found in many regional scale projects.
In Asia, the Asian Association of Remote Sensing (AARS) is generating various
land cover data sets based on (un)supervised classification of multispectral satellite data. In the Africover project (by FAO), image classification techniques are
being used to establish a pan-African land cover data set. The European Commission requires national governments to verify the claims of farmers related
to subsidized crops. These national governments employ companies to make a
first inventory, using image classification techniques, which is followed by field
checks.
Image classification is based on the different spectral characteristics of different materials on the Earths surface, introduced in Section 2.4. This chapter
focusses on classification of multispectral image data. Section 12.2 explains the
concepts of image space and feature space. Image classification is a process that

first

last

back

exit

zoom

contents

index

about

343

12.1. Introduction
operates in the feature space. Section 12.3 gives an overview of the classification
process, the steps involved and the choices to be made. The result of an image
classification needs to be validated to assess its accuracy (Section 12.4). Finally,
two major problems in image classification are addressed in Section 12.5

first

last

back

exit

zoom

contents

index

about

344

12.2. Principle of image classification

12.2

first

Principle of image classification

last

back

exit

zoom

contents

index

about

345

12.2. Principle of image classification

12.2.1 Image space

A digital image is a 2D-array of elements. In each element the energy reflected or
emitted from the corresponding area on the Earths surface is stored. The spatial
arrangement of the measurements defines the image or image space. Depending
on the sensor, data are recorded in n bands (Figure 3.11, repeated here as Figure 12.1). Digital image elements are usually stored as 8-bit DN-values (range:
0255).
Columns

Rows
45

band 3

26 81
53 35 57

Figure 12.1: The structure of a multi-band image

band 1

DN-values

Single pixel

first

band 2

last

back

exit

zoom

contents

index

about

346

12.2. Principle of image classification

12.2.2 Feature space

In one pixel, the values in (for example) two bands can be regarded as components of a two-dimensional vector, the feature vector. An example of a feature
vector is (13, 55), which tells that 13 DN and 55 DN are stored for band 1 and
band 2 respectively. This vector can be plotted in a two-dimensional graph.
Image

band 1
band 2

Image

band 1
band 2
band 3
(v1, v2, v3)

(v1, v2)
band 3
band 2

3-dimensional
Feature Space

2-dimensional
Feature Space

Figure 12.2: Plotting of

the values of a pixel in
the feature space for a two
and three band image.

v1
band 1

band 1

band 2

Similarly, this approach can be visualized for a three band situation in a threedimensional graph. A graph that shows the values of the feature vectors is called
a feature space or feature space plot. Figure 12.2 illustrates how a feature vector
(related to one pixel) is plotted in the feature space for two and three bands.
Usually we only find two axis feature space plots.
Note that plotting values is difficult for a four- or more-dimensional case. A
practical solution when dealing with four or more bands is that all the possible
first

last

back

exit

zoom

contents

index

about

347

12.2. Principle of image classification

Figure 12.3: Scatterplot

of two bands of a digital
image. Note the units (DNvalues) along the x- and yaxes. The intensity at a
point in the feature space
is related to the number of
pixels at that point.

combinations of two bands are plotted separately. For four bands, this already
yields six combinations: bands 1 and 2, 1 and 3, 1 and 4, bands 2 and 3, 2 and 4,
and bands 3 and 4.
Plotting the combinations of the values of all the pixels of one image yields a
large cluster of points. Such a plot is also referred to as a scatterplot (Figure 12.3).
A scatterplot provides information about the combinations of pixel values that
occur within the image. Note that some combinations will occur more frequently
and can be visualized by using intensity or colour.

first

last

back

exit

zoom

contents

index

about

348

12.2. Principle of image classification

Distances and clusters in the feature space

band y (units of 5 DN)

Distance in the feature space is expressed as Euclidian distance and the units
are DN (as this is the unit of the axes). In a two-dimensional feature space the
distance can be calculated according to Pythagoras theorem. In the situation of
Figure 12.4, the distance between (10, 10) and (40, 30) equals the square root of
(40 10)2 + (30 10)2 . For three or more dimensions, the distance is calculated
in a similar way.

(0,0)

first

last

Figure 12.4:
Euclidian
distance between the two
points is calculated using
Pythagoras theorem

band x (units of 5 DN)

back

exit

zoom

contents

index

about

349

12.2. Principle of image classification

12.2.3 Image classification

The scatterplot shown in Figure 12.3 gives information about the distribution
of corresponding pixel values in two bands of an image. Figure 12.5 shows a
feature space in which the feature vectors have been plotted for six specific land
cover classes (grass, water, trees, etc). Each cluster of feature vectors (class) occupies its own area in the feature space. Figure 12.5 shows the basic assumption
for image classification: a specific part of the feature space corresponds to a specific class. Once the classes have been defined in the feature space, each image
pixel can be compared to these classes and assigned to the corresponding class.
Classes to be distinguished in an image classification need to have different
spectral characteristics. This can, for example, be analyzed by comparing spectral reflectance curves (Section 2.4). Figure 12.5 also illustrates the limitation of
image classification: if classes do not have distinct clusters in the feature space,
image classification can only give results to a certain level of reliability.
The principle of image classification is that a pixel is assigned to a class based
on its feature vector, by comparing it to predefined clusters in the feature space.
Doing so for all image pixels results in a classified image. The crux of image
classification is in comparing it to predefined clusters, which requires definition of
the clusters and methods for comparison. Definition of the clusters is an interactive process and is carried out during the training process. Comparison of the
individual pixels with the clusters takes place using classifier algorithms. Both are
explained in the next section.

first

last

back

exit

zoom

contents

index

about

350

12.2. Principle of image classification

255

grass
water
band y

trees
houses
bare soil

Figure 12.5:
Feature
space showing the respective clusters of six
classes; note that each
class occupies a limited
area in the feature space.

wheat

band x

first

last

back

255

exit

zoom

contents

index

about

351

12.3. Image classification process

12.3

Image classification process

The process of image classification (Figure 12.6) typically involves five steps:
1. Selection and preparation of the image data. Depending on the cover types
to be classified, the most appropriate sensor, the most appropriate date(s)
of acquisition and the most appropriate wavelength bands should be selected (Section 12.3.1).
Training
data

Remote
Sensing
data

Training and selection

of algorithm

running the
actual
classification

Figure 12.6: The classification process; most

important component is
the training in combination
with selection of the algorithm.

Classification
data

2. Definition of the clusters in the feature space. Here two approaches are
possible: supervised classification and unsupervised classification. In a supervised classification, the operator defines the clusters during the training
process (Section 12.3.2); in an unsupervised classification a clustering algorithm automatically finds and defines a number of clusters in the feature
space (Section 12.3.3).
3. Selection of classification algorithm. Once the spectral classes have been
defined in the feature space, the operator needs to decide on how the pixels
(based on their DN-values) are assigned to the classes. The assignment can
be based on different criteria (Section 12.3.4).
first

last

back

exit

zoom

contents

index

about

352

12.3. Image classification process

4. Running the actual classification. Once the training data have been established and the classifier algorithm selected, the actual classification can be
carried out. This means that, based on its DN-values, each individual pixel
in the image is assigned to one of the defined classes (Figure 12.7).
5. Validation of the result. Once the classified image has been produced its
quality is assessed by comparing it to reference data (ground truth). This
requires selection of a sampling technique, generation of an error matrix,
and the calculation of error parameters (Section 12.4).

(a)

Figure 12.7: The result

of classification of a multispectral image (a) is a
raster in which each cell
is assigned to some thematic class (b).

(b)

The above points are elaborated in the next sections. Most examples deal
with a two-dimensional situation (two bands) for reasons of simplicity and visualization. In principle, however, image classification can be carried out on any
n-dimensional data set. Visual image interpretation limits itself to an image that
is composed of a maximum of three bands.

12.3. Image classification process

12.3.4 Classification algorithms

After the training sample sets have been defined, classification of the image can
be carried out by applying a classification algorithm. Several classification algorithms exist. The choice of the algorithm depends on the purpose of the classification and the characteristics of the image and training data. The operator
needs to decide if a reject or unknown class is allowed. In the following, three
classifier algorithms are explained. First the box classifier is explained, for its simplicity to help you understanding the principle. In practice, the box classifier is
hardly ever used. In practice the Minimum Distance to Mean and the Maximum
Likelihood classifiers are used.

first

last

back

exit

zoom

contents

index

about

360

12.3. Image classification process

about

362

12.3. Image classification process

Band 2

255

Mean vectors
255

0
Band 2

Band 1

255
Distance
Threshold
0

Band 1

255

first

Figure 12.10:
Principle
of the minimum distance
to mean classification in
a two-dimensional situation. The decision boundaries are shown for a
situation without threshold distance (upper right)
and with threshold distance (lower right).

Unknown

Band 2

255

last

back

exit

Band 1

zoom

contents

255

index

about

363

12.3. Image classification process

Maximum Likelihood classifier
The Maximum Likelihood (ML) classifier considers not only the cluster centre
but also its shape, size and orientation. This is achieved by calculating a statistical distance based on the mean values and covariance matrix of the clusters.
The statistical distance is a probability value: the probability that observation x
belongs to specific cluster. The pixel is assigned to the class (cluster) to which
it has the highest probability. The assumption of most ML classifiers is that the
statistics of the clusters have a normal (Gaussian) distribution.
For each cluster, so-called equiprobability contours can be drawn around
the centres of the clusters. Maximum likelihood also allows the operator to define a threshold distance by defining a maximum probability value. A small
ellipse centred on the mean defines the values with the highest probability of
membership of a class. Progressively larger ellipses surrounding the centre represent contours of probability of membership to a class, with the probability
decreasing away from the centre. Figure 12.11 shows the decision boundaries
for a situation with and without threshold distance.

first

last

back

exit

zoom

contents

index

about

364

12.3. Image classification process

Band 2

255

Mean vectors and variancecovariance matrices

255

Band 2

Band 1

255

Band 1

255

Unknown

Figure 12.11:
Principle
of the maximum likelihood classification. The
decision boundaries are
shown for a situation without threshold distance (upper right) and with threshold distance (lower right).

Band 2

255

Feature space partitioning

Maximum Likelihood Classifier
0

first

last

back

12.5

Problems in image classification

Image classification is a powerful technique to derive thematic classes from

multi-band image data. However, it has certain limitations that you should be
aware of. The most important constraints of pixel based image classification are
that it results in (i) spectral classes, and that (ii) each pixel is assigned to one class
only.
Spectral classes are classes that are directly linked to the spectral bands used
in the classification. In turn, these are linked to surface characteristics. In that
respect one can say that spectral classes correspond with land cover classes (see
also Section 11.3.3). In the classification process a spectral class may be represented by several training classes. Among others this is due to the variability
within a spectral class. Consider a class such as grass; there are different types
of grass, which have different spectral characteristics. Furthermore, the same
type of grass may have different spectral characteristics when considered over
larger areas due to, for example, different soil and climate conditions. A related
topic is that sometimes one is interested in land use classes rather than land
cover classes. Sometimes, a land use class may be comprised of several land
cover classes. Table 12.2 gives some examples of linking spectral land cover and
land use classes. Note that between two columns there can be 1-to-1, 1-to-n,
and n-to-1 relationships. The 1-to-n relationships are a serious problem and can
only be solved by adding data and/or knowledge to the classification procedure.
The data added can be other remote sensing image data (other bands, other moments) or existing spatial data, such as topographic maps, historical land inventories, road maps, et cetera. Usually this is done in combination with adding
expertise knowledge to the process. An example is using historical land cover
data and defining the probability of certain land cover changes. Another exam-

first

last

back

exit

zoom

contents

index

about

CAUTION

369

12.5. Problems in image classification

ple is to use elevation, slope and aspect information. This will prove especially
useful in mountainous regions where elevation differences play an important
role in variations in surface cover types.
Spectral Class
water
grass1
grass2
grass3
bare soil
trees1
trees2
trees3

Land Cover Class

water
grass
grass
grass
bare soil
forest
forest
forest

Table 12.2:
Spectral
classes
distinguished
during classification can
be aggregated to land
cover classes. 1-to-n and
n-to-1 relationships can
exist between land cover
and land use classes.

Land Use Class

shrimp cultivation
nature reserve
nature reserve
nature reserve
nature reserve
nature reserve
production forest
city park

The other main problem and limitation of image classification is that each
pixel is only assigned to one class. When dealing with (relatively) small pixels,
this is not a problem. However when dealing with (relatively) large pixels more
land cover classes will occur within this pixel. As a result, the spectral value
of the pixel is an average of the reflectance of the land cover present within the
pixel. In a standard classification these contributions cannot be traced back and
the pixel will be assigned to one of either classes or even to another class. This
phenomenon is usually referred to as the mixed pixel, or mixel (Figure 12.12). This
problem of mixed pixels is inherent to image classification: assigning the pixel
to one thematic class. The solution to this is to use a different approach, for
example, assigning the pixel to more than one class. This brief introduction into
the problem of mixed pixels also highlights the importance of using data with
the appropriate spatial resolution.
first

last

4. Image classification is sometimes referred to as automatic classification. Do

you agree or disagree? (give argumentation)
5. Draw a simple error matrix, indicate what is on the axes and explain how
the overall accuracy is calculated.

first

last

back

exit

zoom

contents

index

about

Bibliography
[1] Stan Aronoff. Geographic Information Systems: A Management Perspective.
WDL Publications, Ottawa, 1989.
[2] Henk J. Buiten and Jan G. P. W. Clevers. Land Observation by Remote Sensing: Theory and Applications, volume 3 of Current Topics in Remote Sensing.
Gorden & Breach, 1993. 30
[3] Peter A. Burrough and Andrew U. Frank. Geographic Objects with Indeterminate Boundaries. GISDATA Series. Taylor & Francis, London, 1996. 331
[4] Peter A. Burrough and R. McDonnell. Principles of Geographical Information
Systems. Oxford University Press, Oxford, 1998.
[5] European Community. CORINE Land Cover Technical Guide. ECSCEEC
EAEC, Brussels, Belgium, 1993. EUR 12585 EN. 325, 330
[6] Rolf A. de By, editor. Principles of Geographic Information Systems, volume 1 of
ITC Educational Textbook Series. International Institute for Aerospace Survey
and Earth Sciences, Enschede, second edition, 2001. 18, 32, 254

first

last

back

exit

zoom

contents

index

about

375

376

Bibliography
[7] J. A. Deckers, F. O. Nachtergaele, and O. C. Spaargaren, editors. World
reference base for soil resources: introduction. ACCO, Leuven, Belgium, 1998.
318
[8] G. Edwards and K. E. Lowell. Modeling uncertainty in photo-interpretated
boundaries. Photogrammetric Engineering and Remote Sensing, 60(4):337391,
1996. 334
[9] P. Gunn. Airborne magnetic and radiometric surveys. Journal of Australian
Geology and Geophysics, 17:1216, 1997. Special Issue. 220
[10] John Horn. Aerial Photography. ITC Lecture Notes PHM.80. ITC, Enschede,
The Netherlands, 2000. 119
[11] ITC. ITC Textbook of Photo-interpretation. Four volumes, 19631974. 18
[12] ITC. ITC Textbook of Photogrammetry. Five volumes, 19631974. 18
[13] Janes. Janes Space Directory 19971998. Alexandria, Janes Information
Group, 13th edition, 1997. 84
[14] J. H. Kramer. Observation of the Earth and its Environment: Survey of Mission
and Sensors. Springer Verlag, Berlin, Germany, third edition, 1996. 84
[15] Robert Laurini and Derek Thompson. Fundamentals of Spatial Information
Systems, volume 37 of The APIC Series. Academic Press, London, 1992.
[16] Thomas M. Lillesand and Ralph W. Kiefer. Remote Sensing and Image Interpretation. John Wiley & Sons, New York, NY, third edition, 1994. 30, 58,
119, 170

first

last

back

exit

zoom

contents

index

about

377

Bibliography
[17] Keith R. McCloy. Resource Management Information Systems. Taylor & Francis, London, U.K., 1995. 75, 76, 119
[18] Hans Middelkoop. Uncertainty in a GIS, a test for quantifying interpretation output. ITC Journal, 1990(3):225232, 1990. 335
[19] Martien Molenaar. An Introduction to the Theory of Spatial Object Modelling.
Research Monographs in GIS Series. Taylor & Francis, London, 1998. 331
[20] V. Perdigao and A. Annoni. Technical and Methodological Guide for Updating CORINE Land Cover Data Base. EC-JRC, EEA, Brussels, Belgium, 1997.
EUR 17288 EN. 325
[21] Donna J. Peuquet and D. F. Marble, editors. Introductory Readings in Geographic Information Systems. Taylor & Francis, London, 1990.
[22] Colin V. Reeves. Continental scale and global scale geophysical anomaly
mapping. ITC Journal, 1998(2):9198, 1998. 220
[23] David G. Rossiter. Lecture Notes: Methodology for Soil Resource Inventories.
ITC Lecture Notes SOL.27. ITC, Enschede, The Netherlands, 2nd revised
edition, 2000. 318
[24] F. F. Sabins. Remote Sensing: Principles and Interpretation. Freeman & Co.,
New York, NY, third edition, 1996. 30
[25] Toni Schenk. Digital Photogrammetry, volume 1. TerraScience, Laurelville,
1999. 254

first

last

back

exit

zoom

contents

index

about

378

Bibliography
[26] W. Smith and D. Sandwell. Measured and estimated seafloor topography,
version 4.2. Poster RP1, 1997. World Data Center for Marine Geology and
Geophysics. 224
[27] Michael F. Worboys. GIS: A Computing Perspective. Taylor & Francis, London, U.K., 1995.
[28] Alfred J. Zinck. Physiography & Soils. ITC Lecture Notes SOL.41. ITC, Enschede, The Netherlands, 1988. 318

first

last

back

exit

zoom

contents

index

about

Glossary

first

last

back

exit

zoom

contents

index

about

379

380

Glossary

A
Absorption The process in which electromagnetic energy is converted in an
object into other forms of energy (e.g., heat).
Active sensor Sensor with a built in source of energy. The sensor both emits
and receives energy (e.g., radar and laser).
Additive colours The additive principle of colours is based on the three primary colours of light: red, green, blue. All three primary colours together produce white. Additive colour mixing is used, for example,
on computer screens and television.

first

last

back

exit

zoom

contents

index

about

381

Glossary

B
Backscatter The microwave signal reflected by elements of an illuminated
surface in the direction of the radar antenna.
Band

first

Usually related to wavelength band, which indicates a specific range

of the electromagnetic spectrum to which the sensor is sensitive. In
general, it can indicate any layer of an n-dimensional image.

last

back

exit

zoom

contents

index

about

382

Glossary

C
Charge coupled device (CCD) Semi-conductor elements usually aligned as a
linear (scanner) or surface array (video, digital camera). CCDs produce image data.
Class

Classes, usually defined in visual image interpretation and image

classification, are a variable of the nominal type. Class schemes are
applied to describe hierarchical relationships between them.

Cluster

Used in the context of image classification to indicate a concentration

of observations (points in the feature space) related to a training class.

Colour

Colloquial term to indicate hue (IHS space).

Colour film Also known as true colour film used in (aerial) photography. The
principle of colour film is to add sensitized dyes to the silver halide.
Magenta, yellow and cyan dyes are sensitive to red, green and blue
light respectively.
Colour infrared film Film with specific sensitivity for infrared wavelengths.
Typically used in surveys of vegetation.

first

last

back

exit

zoom

contents

index

about

383

Glossary

D
Di-electric constant Parameter that describes the electrical properties of a medium. Reflectivity of a surface and penetration of microwaves into the
material are determined by this parameter.
Digital Elevation Model (DEM) Special case of a DTM. A DEM stores terrain
elevation (surface height) by means of a raster. Elevation refers to a
height expressed with respect to a specific reference.
Digital Terrain Model (DTM) Term indicating a digital description of the terrain relief. A DTM can be stored in different manners (contour lines,
TIN, raster) and may also contain semantic, relief-related information
(breaklines, saddlepoints).

first

last

back

exit

zoom

contents

index

about

384

Glossary

E
Earth Observation (EO) Term indicating the collection of remote sensing techniques performed from space.
Electromagnetic energy Energy with both electric and magnetic components.
Both the wave model and photon model are used to explain this phenomenon. The measurement of reflected and emitted electromagnetic
energy is an essential aspect in remote sensing.
Electromagnetic spectrum The complete range of all wavelengths, from gamma rays (1012 m) up to very long radio waves (1012 m).
Emission

Radiation of electromagnetic energy. Each object with a temperature

above 0 K (-273 C) emits electromagnetic energy.

Emissivity The radiant energy of an object compared to the energy of a blackbody of the same temperature, expressed as a ratio.
Error matrix Matrix that compares samples taken from the source to be evaluated with observations that are considered as correct (reference). The
error matrix allows calculation of quality parameters such as overall
accuracy, error of omission and error of commission.

first

last

back

exit

zoom

contents

index

about

385

Glossary

F
False colour infrared film see Colour infrared film.
Feature space The mathematical space describing the combinations of observations (DN values in the different bands) of a multispectral or multiband image. A single observation is defined by a feature vector.
Feature space plot A two- or three-dimensional graph in which the observations made in different bands are plotted against each other.
Field of view (FOV) The total swath as observed by a sensor-platform system.
Sometimes referred to as total field of view. It can be expressed as an
angle or by the absolute value of the width of the observation.
Filter

first

(1) Physical product made out of glass and used in remote sensing
devices to block certain wavelenghts, e.g., ultraviolet-filter. (2) Mathematical operator used in image processing for modifying the signal,
e.g., a smoothing filter.

last

back

exit

zoom

contents

index

about

386

Glossary

G
Geo-spatial data Data that includes positions in the geographic space. In this
book, usually abbreviated to spatial data.
Geocoding Process of transforming and resampling image data in such way
that these can be used simultaneously with data that are in a specific
map projection. Input for a geocoding process are image data and
control points, output is a geocoded image. A specific category of
geocoded images are orthophotos and orthoimages.
Geographic information Information derived from spatial data, and in the
context of this book, from image data. Information is what is relevant in a certain application context.
Geographic Information System (GIS) A software package that accommodates
the capture, analysis, manipulation and presentation of georeferenced
data. It is a generic tool applicable to many different types of use (GIS
applications).
Georeferencing Process of relating an image to a specific map projection. As a
result, vector data stored in this projection can for example be superimposed on the image. Input for a georeferencing process are image
data and coordinates of ground control points, output is a georeferenced image.
Global Navigation Satellite System (GNSS) The Global Navigation Satellite
System is a global infrastructure for the provision of positioning and
timing information. It consists of the American GPS and Russian
Glonass systems. There is also a proposed European Galileo system.
first

last

back

exit

zoom

contents

index

about

387

Glossary
Ground control points (GCPs) Points which are used to define or validate
a geometric transformation process. Sometimes also referred to as
Ground Control Points stating these have been measured on the ground.
Ground control points should be recognizable both in the image and
in the real world.
Ground range Range direction of the side-looking radar image as projected
onto the horizontal reference plane.
Ground truth A term that may include different types of observations and
measurements performed in the field. The name is imprecise because
it suggests that these are 100% accurate and reliable, and this may be
difficult to achieve.

first

last

back

exit

zoom

contents

index

about

388

Glossary

H
Histogram Tabular or graphical representation showing the (absolute and/or
relative) frequency. In the context of image data it relates to the distribution of the (DN) values of a set of pixels.
Histogram equalization Process used in the visualization of image data to optimize the overall image contrast. Based on the histogram, all available grey levels or colours are distributed in such way that all occur
with equal frequency in the result.

first

last

back

exit

zoom

contents

index

about

389

Glossary

I
Image

A digital file comprising pixels that represent measured local reflectance

(emission or backscatter) values in some designated part of the electromagnetic spectrum. Typically, images are stored in a row-column
system. An image may comprise any number of bands (or channels).
After the reflectance values have been translated into some thematic
variable, the image becomes a raster. With an image, we talk of its
constituent pixels; with a raster we talk of its cells.

Image classification Image classification is the process of assigning pixels to

nominal, i.e., thematic, classes. Input is a multi-band image, output is
a raster in which each cell has a (thematic) code. Image classification
can be realized using a supervised or unsupervised approach.
Image enhancement The process of improving the visual representation of an
image, for example by histogram equalization or using filters.
Image interpretation The process of information extraction from image data.
What is information is defined by the application context. Visual interpretation and pattern recognition techniques can be used.
Image processing system A computer system that is specifically designed to
process image data and to extract information from it by visualizing
the data, application of models and pattern recognition techniques.
Image space The mathematical space describing the (relative) positions of the
observations. Image positions are expressed by their row and column
index.

first

last

back

exit

zoom

contents

index

about

390

Glossary
Incidence angle Angle between the line of sight from the sensor to an element
of an imaged scene and a vertical direction to the scene. One must
distinguish between the nominal incidence angle determined by the
geometry of the radar and the Earths geoidal surface and the local
incidence angle, which takes into account the mean slope of the pixel
in the image.
Infrared waves Electromagnetic radiation in the infrared region of the electromagnetic spectrum. Near-infrared (7001200 nm), middle infrared
(12002500 nm) and thermal infrared (814 m) are distinguished.
Instantaneous field of view (IFOV) The area observed on the ground by a sensor, which can be expressed by an angle or in ground surface units.
Interferometry Computational process that makes use of the interference of
two coherent waves. In the case of imaging radar, two different paths
for imaging cause phase differences from which an interferogram can
be derived. In SAR applications, interferometry is used for constructing a DEM.
Interpretation elements The elements used by the human vision system to interprete a picture or image. Interpretation elements are: tone, texture,
shape, size, pattern, site, association and resolution.

first

last

back

exit

zoom

contents

index

about

391

Glossary

L
Latent image When exposed to light, the silver halide crystals within the photographic emulsion undergo a chemical reaction, which results in an
invisible latent image. The latent is transformed into a visible image by the development process in which the exposed silver halide is
converted into silver grains that appear black.
Look angle The angle of viewing relative to the vertical (nadir) as perceived
from the sensor.

first

last

back

exit

zoom

contents

index

about

392

Glossary

M
Microwaves Electromagnetic radiation in the microwave window, which ranges from 1100 cm.
Mixel

Acronym for mixed pixel. Mixel is used in the context of image classification where different spectral classes occur within the area covered
by one pixel.

Monoplotting Process that enables extraction of accurate (x, y) coordinates from

image data by correcting for terrain relief.
Multispectral scanning Remote sensing technique in which the surface is scanned at the same moment in different wavelength bands. This book
distinguishes two types of scanners: pushbroom and wiskbroom scanners.

first

last

back

exit

zoom

contents

index

about

393

Glossary

N
Nadir

first

The point (or line) directly under the platform during acquisition of
image data.

last

back

exit

zoom

contents

index

about

394

Glossary

O
Objects

Objects are real world features and have clearly identifiable geometric characteristics. In a computer environment, objects are modelled
using an object-based approach in contrast to a field-based approach,
which is more suited for continuous phenomena.

Orbit

The path of a satellite through space. Types of orbits used for remote
sensing satellites are, for example, (near) polar and geostationary.

Orientation Refers to relative or absolute position of a sensor and/or platform.

Interior orientation relates to the components within a camera or sensor. Relative orientation refers to (successive) positions of the sensorplatform system. In photogrammetry, absolute orientation refers to
the orientation of the image relative to a spatial reference system.
Orthoimage An orthoimage and orthophoto have been corrected for terrain
relief. As a result, an orthoimage can be used directly in combination
with geocoded data.

first

last

back

exit

zoom

contents

index

about

395

Glossary

P
Panchromatic Indication of one (wide or narrow) spectral band in the visible
and near-infrared part of the electromagnetic spectrum.
Passive sensor Sensor that records energy that is produced by external sources
such as the Sun and the Earth.
Pattern recognition Term for the collection of techniques used to detect and
identify patterns. Patterns can be found in the spatial, spectral and
temporal domains. An example of spectral pattern recognition is image classification; an example of spatial pattern recognition is segmentation.
Photogrammetry The science and techniques of making measurements from
photos or image data. Photogrammetric procedures are required for
accurate measurements from stereo pairs of aerial photos, image data
or radar data.
Photograph Image obtained by using a camera. The camera produces a negative film, which is can be printed into positive paper product.
Pixel

Acronym for picture element, which is the elementary unit of image

data. The ground pixel size of image data is related to the spatial
resolution of a sensor system it was produced by.

Pixel value The representation of the energy measured at a point, usually expressed as a Digital Number (DN-) value.

first

last

back

exit

zoom

contents

index

about

396

Glossary

Q
Quantization The number of discrete levels applied to store the energy as measured by a sensor, e.g., 8-bit quantization allows 256 levels of energy.

first

last

back

exit

zoom

contents

index

about

397

Glossary

R
Acronym for Radio Detection And Ranging. Radars are active sensors
at wavelengths between 1100 cm.

Radar

Radar equation Mathematical expression that describes the average received

signal level compared to the additive noise level in terms of system
parameters. Principal parameters include the transmitted power, antenna gain,radar cross section,wavelength and range.
The energy (flux) leaving an element of the surface.

Radiance

Radiometric resolution

The smallest observable difference in energy.

RAR

Acronym for Real Aperture Radar.

Raster

A regularly spaced set of cells with associated (field) values. In contrast to a grid, the associated values represent cell values, not point
values. This means that the value for a cell is assumed to be valid for
all locations within the cell.

Reflectance The ratio of the reflected radiation to the total irradiation. Reflectance depends on the wavelength.
Reflection

The process of scattering of of electromagnetic waves by an object.

Relief displacement A shift of the position of an imaged object in the image

due to the local elevation of the object. The shift (magnitude and
direction) depends on sensor-platform characteristics and on the elevation of the object.

first

last

back

exit

zoom

contents

index

about

398

Glossary
Remote sensing (RS) Remote sensing is the instrumentation, techniques and
methods to observe the Earths surface at a distance and to interpret
the images or numerical values obtained in order to acquire meaningful information of particular objects on Earth.
Resampling Process to generate a raster with another orientation and/or different cell size and to assign DN-values using one of the following
methods,nearest neighbour selection, bilinear interpolation and cubic convolution.
Resolution Indicates the smallest observable (measurable) difference at which
objects can still be distinguished. In remote sensing context used in
spatial, spectral and radiometric resolution.
RMS error Root Mean Squared error. A statistical measure of accuracy, similar
to standard deviation, indicating the spread of the measured values
around the true value.

first

last

back

exit

zoom

contents

index

about

399

Glossary

S
SAR

Acronym for Synthetic Aperture Radar. The (high) azimuth resolution

(direction of the flight line) is achieved through off-line processing.
The SAR is able to function as if it has a large virtual antenna aperture, synthesized from many observations with the (relative) small
real antenna of the SAR system.

Scale

The ratio of the distance measured on a (printed) map or image to

that measured on the ground surface between the two same points.
It is expressed as a ratio, e.g.: 1 : 10,000. The 10,000 is referred to as
scale factor.

Scanner

(1) remote sensing sensor that is based on the scanning principle, e.g.,
a multispectral scanner (2) office device to convert analogue products
(photo, map) into digital raster format.

Slant range Image direction as measured along the sequence of line of sight
rays from the radar to each reflecting point in the scene.
Spatial data In the broad sense, spatial data is any data with which position is
associated.
Spatial resolution
Speckle

See resolution.

Interference of backscattered waves stored in the cells of a radar image. It causes the return signals to be extinguished or amplified resulting in random dark and bright pixels in the image.

Spectral band The range of wavelengths to which a channel (single band) of

a multispectral scanner is sensitive.
first

last

back

exit

zoom

contents

index

about

400

Glossary
Spectral resolution

See resolution.

Specular reflection

Mirror-like reflection; bounced-off radiation.

Stereo model A 3D relief model observed through stereoscopic vision of a

stereo pair.
Stereo pair A pair of overlapping photos or images that (partially) cover the
same area from a different position. When appropriately taken, stereo
pairs form a stereo model that can be used for stereoscopic vision and
stereoplotting.
Stereoplotting Process that allows to measure accurate (x, y, z) coordinates
from stereo models.
Stereoscopic vision The ability to perceive distance or depth by observation
with both eyes. In remote sensing, stereoscopic vision is used for the
three-dimensional observation of two images (photos) that are made
from different positions. Stereoscopy is used in visual image interpretation and stereoscopic measurements (stereoplotting). It can yield
3D coordinates.
Subtractive colours The subtractive principle of colours is based on the three
printing colours: cyan, magenta and yellow. All printed colours can
be produced by a combination of these three colours. The subtractive
principle is also used in colour photography.
Sun-synchronous Used to indicate a satellite orbit that is designed in such
way that the satellite always passes the same location on Earth at the
same local time.

first

last

back

exit

zoom

contents

index

about

401

Glossary

T
Training stage Part of the image classification process in which pixels representative for a certain class are identified. Training results in a training set that comprises the statistical characteristics (signatures) of the
classes of interest.
Transmittance The ratio of the radiation transmitted to the total irradiation.

first

last

back

exit

zoom

contents

index

about

402

Glossary

V
Variable, interval A variable that is measured on a continuous scale, but with
no natural zero. It cannot be used to form ratios.
Variable, nominal A variable that is organized in classes, with no natural order, i.e., cannot be ranked.
Variable, ordinal A variable that is organized in classes with a natural order,
and so it can be ranked.
Variable, ratio A variable that is measured on a continuous scale, and with a
natural zero, so can be used to form ratios.
Viewing angle
sor.

first

Angle of observation referring to the vertical (nadir) of the sen-

last

back

exit

zoom

contents

index

about

403

Glossary

W
Wavelength Minimum distance between two events of a recurring feature in
a periodic sequence such as the crests of a wave. Wavelength is expressed as a distance (e.g., m or nm).

first

last

back

exit

zoom

contents

index

about

Index
absorptance, 56
active sensor, 60
radar, 185
additive colours, 282
aerial camera, 111, 122
digital, 148
aerial photography
oblique, 119
vertical, 119
aerospace surveying, 30
altitude, 104
angle
inclination, 104
atmospheric window, 63

maximum likelihood, 363

minimum distance to mean, 361
colour, 277
hue, 284
IHS system, 283
RGB system, 281
spaces, 280
YMC system, 285
coordinate system, 258
dead ground effect, 141
digital photogrammetry, 268
digitizing, 267
DTM, 98, 99, 268
Earth observation, 31
electromagnetic energy, 53
electromagnetic spectrum, 59
optical wavelengths, 58
emissivity, 56
error of commission, 367

blackbody, 56
charged coupled device, 149
charged coupled devices, 160
classification algorithm
box, 360
first

last

back

exit

zoom

contents

index

about

404

405

Index
error of omission, 366

georeferencing, 259
GIS, 253
ground control point, 260, 270
ground observations, 34
ground pixel size, 108
ground truth, 332
ground-based observations, 28

false colour composite, 293

feature space, 346
fiducial marks, 126, 269
field observations, 331
field of view, 163
film, 128
emulsion, 131
scanning, 138
speed, 130
true colour, 134
filter
kernel, 297
optical, 124
filter operations, 297
averaging, 299
edge enhancement, 300
filtering, 208
focal length, 123, 139, 140
frequency, 54

height displacement, 256

histogram, 287
cumulative, 289
equalisation, 290
hue, 284, 312
human vision, 310
image, 107
image classification, 314, 342
supervised, 355
unsupervised, 356
image data, 30
image integration, 253
information extraction, 31
instantaneous field of view, 159
interferometry, 254
differential, 215
radar, 215
interpretation, 210
interpretation elements, 312

general sensitivity, 129

geocoding, 262
geometric transformation, 259
first order, 259
residual errors, 260
root mean square error, 261
first

last

back

exit

zoom

contents

index

about

406

Index
land cover, 325
land use, 325, 368
latent image, 132

passive sensor, 60
pattern, 313
photogrammetry, 119, 253
photography
colour infrared, 136
monochrome, 132
photon, 55, 157
pixel, 107
mixed, 369
platform, 101
aircraft, 102
satellite, 104
pocket stereoscope, 315

mapping, 307
mapping unit
minimum, 326
mirror stereoscope, 315
monoplotting, 266
multispectral scanner, 155
IKONOS-OSA, 173
IRS-1D, 172
Landsat-ETM+, 168
Meteosat-VISSR, 165
NOAA-AVHRR, 166
SPOT-HRVIR, 171

quality
geometric transformation, 261
image classification, 365
photo-interpretation, 333
quantization, 108

orbit
geostationary, 105
period, 104
polar, 105
repeat cycle, 104
sun-synchronous, 105
orientation, 269
orthoimage, 268
orthophoto, 268
overall accuracy, 366
overlap, 144
first

radar, 100
azimuth direction, 192
bands, 190
cross section, 212
equation, 187
foreshortening, 202
ground range, 193
ground range resolution, 197
last

back

exit

zoom

contents

index

about

407

Index
imaging, 187
incidence angle, 193
layover, 203
polarisation, 191
range direction, 192
real aperture, 195
slant range, 193
slant range resolution, 196
synthetic aperture, 198
receiving station, 106
reflectance curve, 72
soil, 75
vegetation, 74
water, 76
reflection, 70
diffuse, 71
specular, 71
relief displacement, 255
Remote Sensing, 30
replicability, 333
resampling, 262
resolution
radiometric, 107, 130, 157
spatial, 108, 142, 159
spectral, 107, 131, 158, 162
revisit time, 108

satellite navigation, 102, 147

scale factor, 140
scanner
across-track, 156
along-track, 161
pushbroom, 160
whiskbroom, 156
scattering, 65
Mie, 68
non-selective, 69
Rayleigh, 66
scatterplot, 347
sensor
active, 85
aerial camera, 89, 122
gamma-ray spectrometer, 88, 221
gravimeter, 223
imaging radar, 100
imaging spectrometer, 93
laser scanner, 98
magnetometer, 224
multispectral scanner, 91, 155
passive, 85
radar altimeter, 99
radiometer, 96
thermal scanner, 95
video camera, 90

satellite communication, 106

first

last

back

exit

zoom

contents

index

about

408

Index
spaceborne missions
operational, 111
spatial-temporal characteristics, 353
spatio-temporal phenomena, 110
speckle, 206, 299
spectral band, 158
spectral sensitivity, 128
Stefan-Boltzmanns Law, 56
stereo model, 269
stereogram, 315
stereoplotting, 269
stereoscopic vision, 315
stereoviewing, 269
subtractive colours, 134, 285
superimposition, 259, 270

wavelength, 53
wavelength band, 91, 107

texture, 313
three-dimensional, 253
tie point, 270
tone, 312
transfer function, 290
transmission,absorbtion, 61
two-dimensional, 253
updating, 270
validation, 365
viewing angle, 159

first

last

back

exit

zoom

contents

index

about

Appendix A
SI units & prefixes
Quantity
Length
Time
Temperature
Energy
Power

first

last

back

Table A.1: Relevant SI

units in the context of remote sensing

SI unit
metre (m)
second (s)
kelvin (K)
joule (J)
watt (W) (J/s)

exit

zoom

contents

index

about

409

410

Appendix
Prefix
tera (T)
giga (G)
mega (M)
kilo (k)
centi (c)
milli (m)
micro ()
nano (n)
pico (p)

Unit
centimetre
millimetre
micron
micrometre
nanometre

Parameter
speed of light

C
inch
foot
mile

first

last

back

Table A.2: Unit prefix notation

Multiplier
1012
109
106
103
102
103
106
109
1012

SI Equivalent
102 m
103 m
106 m
106 m
109 m

Table A.3: Common units

of wavelength

Value
2.9979 108 m/s
( C + 273.15) K
2.54 cm
30.48 cm
1, 609 m

Table A.4: Constants and

non-SI units

exit

zoom

contents

index

about

KM 250 Parts
89% (9)
KM 250 Parts
465 pages
Instant ebooks textbook Introduction to Remote Sensing, 6th Edition James B. Campbell download all chapters
50% (2)
Instant ebooks textbook Introduction to Remote Sensing, 6th Edition James B. Campbell download all chapters
23 pages
PrinciplesRemoteSensing Contents
No ratings yet
PrinciplesRemoteSensing Contents
36 pages
Petrel Fundamentals
No ratings yet
Petrel Fundamentals
19 pages
Detailed Lesson Plan in Science 7
100% (4)
Detailed Lesson Plan in Science 7
11 pages
A Seminar Report On: "Wireless Charging"
67% (6)
A Seminar Report On: "Wireless Charging"
36 pages
Principles of Remote Sensing
0% (1)
Principles of Remote Sensing
410 pages
Principles of Remote Sensing
No ratings yet
Principles of Remote Sensing
410 pages
Remote Sensing Book PDF
No ratings yet
Remote Sensing Book PDF
540 pages
Principles Remote Sensing PDF
No ratings yet
Principles Remote Sensing PDF
540 pages
PrinciplesRemoteSensing (Unlocked by WWW - Freemypdf.com) PDF
No ratings yet
PrinciplesRemoteSensing (Unlocked by WWW - Freemypdf.com) PDF
591 pages
PrinciplesRemoteSensing PDF
No ratings yet
PrinciplesRemoteSensing PDF
591 pages
DIP& Storage
No ratings yet
DIP& Storage
14 pages
Guide To Instruments and Methods of Observation: Volume IV - Space-Based Observations
No ratings yet
Guide To Instruments and Methods of Observation: Volume IV - Space-Based Observations
223 pages
2 RS Sensors and Platforms
No ratings yet
2 RS Sensors and Platforms
57 pages
Erdas
No ratings yet
Erdas
770 pages
ERDAS FieldGuide
No ratings yet
ERDAS FieldGuide
770 pages
Hyperspectral Signal Processing To Identify Land Cover Pattern PDF
No ratings yet
Hyperspectral Signal Processing To Identify Land Cover Pattern PDF
78 pages
z01_978-0-19-807239-3_01
No ratings yet
z01_978-0-19-807239-3_01
12 pages
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen - The complete ebook is available for download with one click
100% (1)
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen - The complete ebook is available for download with one click
39 pages
Навчальний посібник - Основи дистанційного зондування - Історія та практика - за 2019рік - compressed
No ratings yet
Навчальний посібник - Основи дистанційного зондування - Історія та практика - за 2019рік - compressed
318 pages
Contents
No ratings yet
Contents
9 pages
1999 Fundamentals of Remote Sensing
No ratings yet
1999 Fundamentals of Remote Sensing
225 pages
remote sensing abstracts
No ratings yet
remote sensing abstracts
8 pages
Envi Sci and Eng'g - Learning Material Topic 4
No ratings yet
Envi Sci and Eng'g - Learning Material Topic 4
33 pages
A Brief Introduction To Remote Sensing
No ratings yet
A Brief Introduction To Remote Sensing
4 pages
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen - The ebook is ready for download, no waiting required
100% (2)
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen - The ebook is ready for download, no waiting required
46 pages
3 Earth Observation Sensors and Platforms
No ratings yet
3 Earth Observation Sensors and Platforms
51 pages
Atmospheric Interactions Imaging Systems Image Processing & Classifications
No ratings yet
Atmospheric Interactions Imaging Systems Image Processing & Classifications
12 pages
KULIAH 2 MRT 10-Perpetaan
No ratings yet
KULIAH 2 MRT 10-Perpetaan
85 pages
EO - Lecture 1 - Introduction 2015
No ratings yet
EO - Lecture 1 - Introduction 2015
35 pages
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen download
100% (1)
Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen download
45 pages
Guide To Instruments and Methods 8 - IV - en - 2021
No ratings yet
Guide To Instruments and Methods 8 - IV - en - 2021
197 pages
An Overview of Remote Sensing and Geodesy For Epidemiology and Public Health Application
No ratings yet
An Overview of Remote Sensing and Geodesy For Epidemiology and Public Health Application
36 pages
Photogrammetry AOC From Previous Session
No ratings yet
Photogrammetry AOC From Previous Session
4 pages
Instant download Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen pdf all chapter
100% (3)
Instant download Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen pdf all chapter
84 pages
Remote_sensing_of_the_environment
No ratings yet
Remote_sensing_of_the_environment
8 pages
Basic Concepts of Remote Sensing
100% (1)
Basic Concepts of Remote Sensing
43 pages
10.4324 9780203761038 Previewpdf
100% (1)
10.4324 9780203761038 Previewpdf
40 pages
Introduction to Remote Sensing Fifth Edition James B. Campbell Phd - Instantly access the complete ebook with just one click
100% (2)
Introduction to Remote Sensing Fifth Edition James B. Campbell Phd - Instantly access the complete ebook with just one click
47 pages
image resolution remote sensing and GIS
No ratings yet
image resolution remote sensing and GIS
38 pages
OverviewRemoteSens 2000
No ratings yet
OverviewRemoteSens 2000
35 pages
Fundamentals of Satellite Remote Sensing
100% (1)
Fundamentals of Satellite Remote Sensing
8 pages
Day 1 Session 3- Mr_ K_R_ Sridhara Introduction to Remote Sensing and GIS (1)
No ratings yet
Day 1 Session 3- Mr_ K_R_ Sridhara Introduction to Remote Sensing and GIS (1)
59 pages
Introduction to Remote Sensing, 6th Edition James B. Campbell pdf download
No ratings yet
Introduction to Remote Sensing, 6th Edition James B. Campbell pdf download
53 pages
Previewpdf
No ratings yet
Previewpdf
1,629 pages
Remote Sensing Notes
100% (1)
Remote Sensing Notes
5 pages
BuildingEarthObservationCameras Contentdetails2
No ratings yet
BuildingEarthObservationCameras Contentdetails2
8 pages
Introduction to Remote Sensing, 6th Edition James B. Campbell - The latest ebook is available, download it today
100% (1)
Introduction to Remote Sensing, 6th Edition James B. Campbell - The latest ebook is available, download it today
74 pages
101-04-24-18-15-4455054747
No ratings yet
101-04-24-18-15-4455054747
164 pages
Download Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen ebook All Chapters PDF
100% (5)
Download Remote Sensing from Air And Space SPIE Press Monograph Vol PM162 R. C. Olsen ebook All Chapters PDF
71 pages
Remote Sensing - UNIT I
No ratings yet
Remote Sensing - UNIT I
25 pages
Fundamentals of Satellite Remote Sensing
No ratings yet
Fundamentals of Satellite Remote Sensing
7 pages
Remote Sensing
No ratings yet
Remote Sensing
49 pages
Geographic Information Systems: GIS Data - Remote Sensing
No ratings yet
Geographic Information Systems: GIS Data - Remote Sensing
38 pages
CPTUtimbunan
No ratings yet
CPTUtimbunan
12 pages
L1-Intro 2
No ratings yet
L1-Intro 2
24 pages
Convolution: 2 Historical Developments
No ratings yet
Convolution: 2 Historical Developments
10 pages
Build A Mode Lfrom Scratch in Petrel - 1
No ratings yet
Build A Mode Lfrom Scratch in Petrel - 1
15 pages
GCDkit Manual
No ratings yet
GCDkit Manual
272 pages
Journal of The International Society of Swimming Coaching: Volume 2, Issue 1 ISSN 1839-3659
100% (1)
Journal of The International Society of Swimming Coaching: Volume 2, Issue 1 ISSN 1839-3659
50 pages
Introduction To Petroleum Geology
No ratings yet
Introduction To Petroleum Geology
6 pages
1 s2.0 S0360131511001412 Main
No ratings yet
1 s2.0 S0360131511001412 Main
10 pages
Working With Spreadsheets in Oasis Montaj
No ratings yet
Working With Spreadsheets in Oasis Montaj
18 pages
1 s2.0 S2214241X15000073 Main
No ratings yet
1 s2.0 S2214241X15000073 Main
9 pages
GM-SYS Profile DC Shift
No ratings yet
GM-SYS Profile DC Shift
1 page
Speed Suits and Swimming T'S: Alyssa J. Ha & Paul M. Sommers
No ratings yet
Speed Suits and Swimming T'S: Alyssa J. Ha & Paul M. Sommers
5 pages
061cardimona Resistivity Overview
No ratings yet
061cardimona Resistivity Overview
10 pages
Eps SCHEDULE
No ratings yet
Eps SCHEDULE
10 pages
Wimax Project Planning PDF
No ratings yet
Wimax Project Planning PDF
29 pages
Terms of Reference - Consultancy To Conduct Short Course Training in Key Skills To APHRC Staff
No ratings yet
Terms of Reference - Consultancy To Conduct Short Course Training in Key Skills To APHRC Staff
5 pages
Revised Capstone Project
No ratings yet
Revised Capstone Project
37 pages
Guardiana, Wyeth Gain Mayrong BM701A Logistics Management Activity
No ratings yet
Guardiana, Wyeth Gain Mayrong BM701A Logistics Management Activity
3 pages
Som Unit. 01 PDF
No ratings yet
Som Unit. 01 PDF
64 pages
Cybersecurity Roadmap For A Beginner
100% (2)
Cybersecurity Roadmap For A Beginner
42 pages
Dedm PDF
No ratings yet
Dedm PDF
115 pages
6 - Acronis Cyber Notary
No ratings yet
6 - Acronis Cyber Notary
23 pages
Eshan Jain Research Project On Merger and Acquisition
No ratings yet
Eshan Jain Research Project On Merger and Acquisition
21 pages
Yokozawa Takafumi No Baai Vol. 1
No ratings yet
Yokozawa Takafumi No Baai Vol. 1
86 pages
Airliner
100% (1)
Airliner
8 pages
Rubrics
No ratings yet
Rubrics
1 page
Modulus of Elasticity11
No ratings yet
Modulus of Elasticity11
4 pages
Comm 324 M5A1 Notes
No ratings yet
Comm 324 M5A1 Notes
4 pages
Class Test
No ratings yet
Class Test
2 pages
Configurar s7 Conection
No ratings yet
Configurar s7 Conection
35 pages
Instant Access to ROUTLEDGE CONCISE GRAMMARS Exacto A Practical Guide to Spanish Grammar 3rd ed 2018 3rd Edition Ane Ortega ebook Full Chapters
100% (7)
Instant Access to ROUTLEDGE CONCISE GRAMMARS Exacto A Practical Guide to Spanish Grammar 3rd ed 2018 3rd Edition Ane Ortega ebook Full Chapters
61 pages
Physical Fitness and Exercise - Health - 7th Grade by Slidesgo
No ratings yet
Physical Fitness and Exercise - Health - 7th Grade by Slidesgo
56 pages
Stochastic Interacting System PDF
No ratings yet
Stochastic Interacting System PDF
346 pages
Paper Making
No ratings yet
Paper Making
2 pages
XII Physics Chapter 6 - Electrictromagnetic Induction Saju Hsslive
No ratings yet
XII Physics Chapter 6 - Electrictromagnetic Induction Saju Hsslive
7 pages
D1- Maxhub v6 classic
No ratings yet
D1- Maxhub v6 classic
2 pages
Palatal Rugoscopy As A Method of Sex Determination in Forensic Science
No ratings yet
Palatal Rugoscopy As A Method of Sex Determination in Forensic Science
3 pages
Step7 - FM354
No ratings yet
Step7 - FM354
364 pages
Voice Authentication App Tutorial
No ratings yet
Voice Authentication App Tutorial
13 pages
Correlation Does Not Imply Causation
No ratings yet
Correlation Does Not Imply Causation
9 pages