THEORY: Remote sensing is the science and art of obtaining information about an

object, area, or phenomenon through the analysis of data acquired by a device that
is not contact with the object, area or phenomenon under investigation. It is the
utilization of satellite or air-based sensors or detectors with the help of EMR
(Electromagnetic Radiation), in other words; propagated signals.

OBJECTIVE
The overall objective of image classification procedures is to automatically categorize
all pixels of an image into land cover classes or themes.
In this process, we will categorize the five major areas of Yangon;

1. Lake
2. River
3. Urban
4. Vegetation
5. Bare Land by supervised method.

Steps for this procedure are as follows:

1. Training Stage

2. Classification Stage

3. Output Stage

Advantages
Remote sensing is used widely today because of these useful facts;

 Systematic data collection

 Obtain information about the inaccessible areas
 Information about the three dimensions of real objects and areas
 Repeatability
 Global converge
 Easier to get basic information about a site area
 Multipurpose information

For supervised classification

Decisions rules can be classified as parametric and non-parametric rules.

The parametric rule has two contents;

1. Minimum distance to mean

2. Maximum likelihood

Maximum likelihood classification assumes that the

statistics for each class in each band are normally
distributed and calculates the probability that a given
pixel belongs to a specific class.

Unless we select a probability threshold, all pixels are

classified.

Each pixel is assigned to the class that has the highest probability (that is, the
maximum likelihood).

If the highest probability is smaller than a threshold we specify, the pixel remains
unclassified.

In wefrewfe rghg Second pitcture shows difference between methods used so it can say we
should use the best or the most appropriate method for our image.

PROCEDURES
Step 1: Open Yangon City map at 1989 using
‘Erdas’ application .
Step 2: On the ‘Raster’ tap toolbar, click the
‘Supervised’ tap, and then choose ‘Signature
Editor’.
Step 3: Select ‘Drawing’ tool bar and choose AOI
shape of polygon.
Then enclose an area of lake, river, urban,
vegetation, bare ground. Add them to the
Signature Editor by giving different colors to
each band.
For individual band, we choose there to four
samples and made average of them in the signature editor.
And these all must be uni-model in histogram.
Step 4: Select Supervised and select the maximum likelihood parametric
decision rules and click the Classify zeros and Use Probabilities. Then click OK
and the Classified Image comes out on the screen of the computer after
entering the ‘New Viewer’.
Step 5: Open Raster, there go to Supervised and
then Accuracy Assessment box comes out. There,
open the Classified image file and then go to Edit
tap and go to ‘Creat/ Random Points’ and put 50
to create 50 Random points for accuracy
assessment.
Step 6: Load points onto the Reference map and
check those 50 random points by visual inspection, adding the corresponding
class value, whether the result shown corresponds to the respective band or
not.
Step 7: Click the Report tap and go to Cell Report in the Accuracy Assessment
box to generate Accuracy Assessment Report. Make Error Matrix, Accuracy
Totals and Kappa Statistics. Classification Accuracy Assessment Report is our
final step.
REFERENCE DATA
Lake- 1, River- 2, Vegetation- 3, Urban- 4, Bare Land- 5
ERROR MATRIX

Classified Data Lake River Vegetation Urban Bare Land Row Total
Lake 8 0 0 0 0 8
River 0 10 0 0 0 10
Vegetation 0 0 10 0 0 10
Urban 2 0 0 10 0 12
Bare Land 0 0 0 0 10 10
Column Total 10 10 10 10 10 50

ACCURACY TOTALS
Class Name Reference Classified Number Producers Users
 Totals Totals Correct Accuracy Accuracy
Lake 10 8 8 80% 100%
River 10 10 10 100% 100%
Vegetation 10 10 10 100% 100%
Urban 10 10 10 100% 100%
Bare Land 10 12 10 100% 83.33%
Totals 50 50 48

Overall Classification Accuracy= 96.00

CONCLUSION
User has the commission error of 16.77% because of error at lake.
CLASSIFICATION ACCURACY ASSESSMENT REPORT
Image File :
c:/users/dell/desktop/rs_practical/image_classification/result
1/supermax - copy.img
User Name : Dell
Date : Mon Mar 11 11:28:32 2019

ERROR MATRIX
-------------
Reference Data
--------------
Classified Data Unclassifi lake river vegetation
--------------- ---------- ---------- ---------- ----------
Unclassified 0 0 0 0
lake 0 8 0 0
river 0 0 10 0
vegetation 0 0 0 10
bareland 0 0 0 0
urban 0 2 0 0
Column Total 0 10 10 10

Reference Data
--------------
Classified Data bareland urban Row Total
--------------- ---------- ---------- ----------
Unclassified 0 0 0
lake 0 0 8
river 0 0 10
vegetation 0 0 10
bareland 10 0 10
urban 0 10 12
Column Total 10 10 50

----- End of Error Matrix -----

ACCURACY TOTALS
----------------
 Class Reference Classified Number Producers Users
Name Totals Totals Correct Accuracy Accuracy
---------- ---------- ---------- ------- --------- -----
Unclassified 0 0 0 --- ---
lake 10 8 8 80.00% 100.00%
river 10 10 10 100.00% 100.00%
vegetation 10 10 10 100.00% 100.00%
bareland 10 10 10 100.00% 100.00%
urban 10 12 10 100.00% 83.33%

Totals 50 50 48

Overall Classification Accuracy = 96.00%

----- End of Accuracy Totals -----

KAPPA (K^) STATISTICS

---------------------

Overall Kappa Statistics = 0.9500

Conditional Kappa for each Category.

------------------------------------

Class Name Kappa

---------- -----
Unclassified 0.0000
lake 1.0000
river 1.0000
vegetation 1.0000
bareland 1.0000
urban 0.7917

----- End of Kappa Statistics -----