2d Computer Vision Principles Algorithms And

Applications Yujin Zhang download

https://ebookbell.com/product/2d-computer-vision-principles-
algorithms-and-applications-yujin-zhang-50902930

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

2d Computer Graphics In Modern C And Standard Library Hkan Blomqvist

https://ebookbell.com/product/2d-computer-graphics-in-modern-c-and-
standard-library-hkan-blomqvist-50378156

Computer Graphics Using Java 2d And 3d Liang Y Daniel Zhang

https://ebookbell.com/product/computer-graphics-using-
java-2d-and-3d-liang-y-daniel-zhang-21995762

Introduction To Computer Graphics Using Java 2d And 3d Frank Klawonn

https://ebookbell.com/product/introduction-to-computer-graphics-using-
java-2d-and-3d-frank-klawonn-2107070

Introduction To Computer Graphics Using Java 2d And 3d 2nd Ed 2012

Klawonn

https://ebookbell.com/product/introduction-to-computer-graphics-using-
java-2d-and-3d-2nd-ed-2012-klawonn-21986742
The Video Games Guide 1000 Arcade Console And Computer Games 19622012
2d Ed Matt Fox

https://ebookbell.com/product/the-video-games-guide-1000-arcade-
console-and-computer-games-19622012-2d-ed-matt-fox-49198118

2d Materials For Surface Plasmon Resonancebased Sensors Sanjeev Kumar

Raghuwanshi

https://ebookbell.com/product/2d-materials-for-surface-plasmon-
resonancebased-sensors-sanjeev-kumar-raghuwanshi-46150346

2d Monoelements Properties And Applications 1st Edition Inamuddin

https://ebookbell.com/product/2d-monoelements-properties-and-
applications-1st-edition-inamuddin-46608926

2d Monoelemental Materials Xenes And Related Technologies Beyond

Graphene Zongyu Huang

https://ebookbell.com/product/2d-monoelemental-materials-xenes-and-
related-technologies-beyond-graphene-zongyu-huang-46703870

2d Nanomaterials Chemistry And Properties Ram K Gupta

https://ebookbell.com/product/2d-nanomaterials-chemistry-and-
properties-ram-k-gupta-47173282
 2D Computer
Vision
Principles, Algorithms
and Applications

12497 9789811245084 tp.indd 1 8/2/22 10:48 AM

 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

 2D
Vision
Computer
Principles, Algorithms
and Applications

YU-JIN ZHANG
Tsinghua University, China

World Scientific

12497 9789811245084 tp.indd 2 8/2/22 10:48 AM

 Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

Library of Congress Cataloging-in-Publication Data

Names: Zhang, Yu-Jin, 1954– author.
Title: 2D computer vision : principles, algorithms and applications /
Yu-Jin Zhang, Tsinghua University, China.
Other titles: Two dimensional computer vision
Description: Hackensack, NJ : World Scientific, 2022. |
Includes bibliographical references and index.
Identifiers: LCCN 2021046188 | ISBN 9789811245084 (hardcover) |
ISBN 9789811245091 (ebook for institutions) | ISBN 9789811245107 (ebook for individuals)
Subjects: LCSH: Computer vision.
Classification: LCC TA1634 .Z46 2022 | DDC 621.39/93--dc23
LC record available at https://lccn.loc.gov/2021046188

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Copyright © 2022 by World Scientific Publishing Co. Pte. Ltd.

All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.

For any available supplementary material, please visit

https://www.worldscientific.com/worldscibooks/10.1142/12497#t=suppl

Desk Editors: Jayanthi Muthuswamy/Steven Patt

Typeset by Stallion Press

Email: [email protected]

Printed in Singapore

Jayanthi - 12497 - 2D Computer Vision.indd 1 9/2/2022 2:15:34 pm

February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page v

Preface

This book is a special textbook that introduces the basic principles,

typical methods, and practical techniques of 2D computer vision. It
can serve as the first computer vision course material for undergrad-
uates of related majors in university and higher-engineering colleges,
and then they can study “3D Computer Vision: Principle, Algorithm,
and Applications”.
This book mainly covers the introductory content of computer
vision from a selection of materials. This book is mainly for infor-
mation majors but also useful for learners with different professional
backgrounds. This book is self-contained in contents and also con-
siders the needs of self-study readers. Readers can not only solve
some specific problems in practical applications but also lay a foun-
dation for further study and research on high-level computer vision
technology.
This book pays more attention to practicality in writing. Con-
sidering that computer vision technology has been involved in many
professional fields (but in which not all people are specialized in com-
puter vision technology) in recent years, so this book does not empha-
size the theoretical system too much, minimizes formula derivation,
and focuses on commonly used methods. This book provides many
examples through intuitive explanation to help readers understand
abstract concepts. A subject index (marked in bold in the text) is
given at the end of the book.
This book provides a large number of self-test questions (including
hints and answers). In terms of purpose, on the one hand, it is conve-
nient for self-study readers to judge whether they have mastered the

v
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page vi

vi 2D Computer Vision: Principles, Algorithms and Applications

key content, and on the other hand, it is also convenient for teachers
to carry out online teaching and strengthen teacher–student inter-
action during teaching. The types of questions are multiple-choice
questions, which can be easily evaluated by a computer. In terms
of content, many questions express the basic concepts in a differ-
ent way and supplement the text so that learners can deepen their
understanding. Some questions list descriptions that are similar but
different or even opposite in meaning. Through dialectical thinking of
pros and cons, learners can also deeply understand the essence. The
hints have been provided for all self-test questions, allowing readers
to obtain more information to further identify the meaning of the
questions. At the same time, each question can be said to be divided
into two levels in this way. Readers can complete the self-test after
reading the hints to show that they basically understand it whereas
to complete the self-test without looking at the hints indicates that
they have an even better grasp.
From a structural point of view, this book has 13 chapters, two
appendices, answers to self-test questions, and subject index. Under
these 17 first-level headings, there are 109 second-level headings (sec-
tions) and 176 third-level headings (sub-sections). The book has a
total of nearly 500,000 words (including figures, drawings, tables, for-
mulas, etc.), a total of 278 numbered pictures, 30 numbered tables,
and 497 numbered formulas. In order to facilitate teaching and under-
standing, this book gives a total of 121 examples of various types and
233 self-test questions (all with hints and answers). In addition, there
are a catalog of more than 100 references and more than 600 terms
used for indexing at the end of the book.
This book generally considers the following three aspects from
the knowledge requirements of the prerequisite courses: (i) mathe-
matics, including linear algebra and matrix theory, as well as basic
knowledge of statistics, probability theory, and random modeling; (ii)
computer science, including the mastery of computer software tech-
nology, understanding of computer structure system, and application
of computer programming methods; (iii) electronics, including on the
one hand, the characteristics and principles of electronic equipment,
and on the other hand, circuit design and other content. In addi-
tion, it is best to study this book after finishing the course on signal
processing.
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page vii

Preface vii

Thanks to the editors of the publisher for carefully com-

posing the manuscript, followed by serious review and attentive
modifications.
Finally, the author thanks his wife Yun He and daughter Heming
Zhang for their understanding and support in all aspects.
 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page ix

About the Author

Yu-Jin Zhang received his Ph.D. degree in

Applied Science from the State University of
Liège, Liège, Belgium, in 1989. From 1989 to
1993, he was a post-doc fellow and research fellow
at the Delft University of Technology, Delft, the
Netherlands. In 1993, he joined the Department
of Electronic Engineering, Tsinghua University,
Beijing, China, where he was a professor (since
1997) and has been a tenured professor (since
2014) of Image Engineering.
He is active in the teaching and research of image engineering,
has opened and taught more than 10 undergraduate and graduate
courses, and published more than 550 research papers and more than
50 books. He has served as the program chair of a number of inter-
national conferences. He is a fellow and the honorary chairman of
supervisors of China Society of Image and Graphics (CSIG). He is
a fellow of International Society for Optical Engineering (SPIE) for
achievements in image engineering.

ix
 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xi

Contents

Preface v
About the Author ix

Chapter 1. Computer Vision Fundamentals 1

1.1 Vision Basis . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Vision . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Visual sensation and visual
perception . . . . . . . . . . . . . . . . . . 3
1.1.3 Visual process . . . . . . . . . . . . . . . . 4
1.2 Vision and Image . . . . . . . . . . . . . . . . . . . 8
1.2.1 Images and digital images . . . . . . . . . 9
1.2.2 Pixel and image representation . . . . . . 11
1.2.3 Image storage and format . . . . . . . . . 12
1.2.4 Image display and printing . . . . . . . . . 18
1.3 Vision Systems and Image Techniques . . . . . . . 21
1.3.1 Vision system flowchart . . . . . . . . . . 22
1.3.2 Three layers of image engineering . . . . . 22
1.3.3 Categories of image technology . . . . . . 24
1.4 Overview of the Structure and Content
of This Book . . . . . . . . . . . . . . . . . . . . . 26
1.4.1 Structural framework and
main content . . . . . . . . . . . . . . . . 26
1.4.2 Overview of each chapter . . . . . . . . . . 27
1.4.3 Prerequisites . . . . . . . . . . . . . . . . . 29

xi
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xii

xii 2D Computer Vision: Principles, Algorithms and Applications

1.5 Key Points and References for Each Section . . . . 30

Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 31
References . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Chapter 2. 2D Image Acquisition 41
2.1 Acquisition Device and Performance Index . . . . 42
2.1.1 CCD sensor . . . . . . . . . . . . . . . . . 43
2.1.2 CMOS sensor . . . . . . . . . . . . . . . . 45
2.1.3 Common performance indicators . . . . . 45
2.1.4 Image acquisition process . . . . . . . . . 46
2.2 Image Brightness Imaging Model . . . . . . . . . . 47
2.2.1 Fundamentals of photometry . . . . . . . . 47
2.2.2 Uniform illuminance . . . . . . . . . . . . 49
2.2.3 A simple brightness imaging model . . . . 52
2.3 Image Space Imaging Model . . . . . . . . . . . . 53
2.3.1 Projection imaging geometry . . . . . . . 53
2.3.2 Basic imaging model . . . . . . . . . . . . 55
2.3.3 General imaging model . . . . . . . . . . . 59
2.4 Sampling and Quantization . . . . . . . . . . . . . 64
2.4.1 Spatial and amplitude resolutions . . . . . 64
2.4.2 Image data volume and quality . . . . . . 66
2.5 Relationship Between Pixels . . . . . . . . . . . . . 71
2.5.1 Pixel neighborhood and connectivity . . . 71
2.5.2 Distance between pixels . . . . . . . . . . 73
2.6 Key Points and References for Each Section . . . . 75
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 76
References . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Chapter 3. Spatial Domain Enhancement 83
3.1 Operation Between Images . . . . . . . . . . . . . 84
3.1.1 Arithmetic operations . . . . . . . . . . . 84
3.1.2 Logical operation . . . . . . . . . . . . . . 88
3.2 Image Gray-scale Mapping . . . . . . . . . . . . . 89
3.2.1 Image negation . . . . . . . . . . . . . . . 89
3.2.2 Contrast stretching . . . . . . . . . . . . . 90
3.2.3 Dynamic range compression . . . . . . . . 91
3.3 Histogram Equalization . . . . . . . . . . . . . . . 92
3.3.1 Image histogram . . . . . . . . . . . . . . 92
3.3.2 Principles and steps . . . . . . . . . . . . . 94
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xiii

Contents xiii

3.4 Histogram Specification . . . . . . . . . . . . . . . 97

3.4.1 Principles and steps . . . . . . . . . . . . . 97
3.4.2 Single mapping law and group mapping
law . . . . . . . . . . . . . . . . . . . . . . 100
3.5 Spatial Domain Convolution Enhancement . . . . 104
3.5.1 Mask convolution . . . . . . . . . . . . . . 104
3.5.2 Spatial filtering . . . . . . . . . . . . . . . 105
3.6 Key Points and References for Each Section . . . . 111
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 113
References . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Chapter 4. Frequency Domain Enhancement 123

4.1 Fourier Transform and Frequency Domain
Enhancement . . . . . . . . . . . . . . . . . . . . . 124
4.1.1 Fourier transform . . . . . . . . . . . . . . 124
4.1.2 Fourier transform properties . . . . . . . . 126
4.1.3 Frequency domain enhancement . . . . . . 128
4.2 Frequency Domain Low-Pass Filter . . . . . . . . . 129
4.2.1 Ideal low-pass filter . . . . . . . . . . . . . 129
4.2.2 Butterworth low-pass filter . . . . . . . . . 131
4.3 Frequency Domain High-Pass Filter . . . . . . . . 133
4.3.1 Ideal high-pass filter . . . . . . . . . . . . 133
4.3.2 Butterworth high-pass filter . . . . . . . . 133
4.4 Band-Pass Filter and Band-Stop Filter . . . . . . 135
4.4.1 Band-pass filter . . . . . . . . . . . . . . . 135
4.4.2 Band-stop filter . . . . . . . . . . . . . . . 137
4.4.3 Relation between band-pass filter and
band-stop filter . . . . . . . . . . . . . . . 138
4.4.4 Notch filter . . . . . . . . . . . . . . . . . 139
4.4.5 Interactively eliminate periodic noise . . . 141
4.5 Homomorphic Filter . . . . . . . . . . . . . . . . . 143
4.5.1 Homomorphic filtering process . . . . . . . 144
4.5.2 Homomorphic filter denoising . . . . . . . 146
4.6 Key Points and References for Each Section . . . . 147
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 149
References . . . . . . . . . . . . . . . . . . . . . . . . . . 157
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xiv

xiv 2D Computer Vision: Principles, Algorithms and Applications

Chapter 5. Image Restoration 159

5.1 Image Degradation and Model . . . . . . . . . . . 160
5.1.1 Image degradation model . . . . . . . . . . 161
5.1.2 Properties of the image degradation
model . . . . . . . . . . . . . . . . . . . . 162
5.2 Inverse Filtering . . . . . . . . . . . . . . . . . . . 164
5.2.1 Unconstrained restoration . . . . . . . . . 164
5.2.2 Inverse filtering model . . . . . . . . . . . 165
5.3 Wiener Filtering . . . . . . . . . . . . . . . . . . . 167
5.3.1 Constrained restoration . . . . . . . . . . 167
5.3.2 Wiener filter . . . . . . . . . . . . . . . . . 167
5.4 Geometric Distortion Correction . . . . . . . . . . 169
5.4.1 Spatial transformation . . . . . . . . . . . 170
5.4.2 Gray-level interpolation . . . . . . . . . . 172
5.5 Image Repairing . . . . . . . . . . . . . . . . . . . 175
5.5.1 Principle of image repairing . . . . . . . . 176
5.5.2 Image repair examples . . . . . . . . . . . 177
5.6 Key Points and References for Each Section . . . . 181
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 183
References . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Chapter 6. Color Enhancement 189
6.1 Color Vision . . . . . . . . . . . . . . . . . . . . . 190
6.1.1 Three primary colors and color
representation . . . . . . . . . . . . . . . . 190
6.1.2 Chromaticity diagram . . . . . . . . . . . 192
6.2 Color Model . . . . . . . . . . . . . . . . . . . . . 195
6.2.1 RGB model . . . . . . . . . . . . . . . . . 195
6.2.2 HSI model . . . . . . . . . . . . . . . . . . 197
6.2.3 Conversion from RGB to HSI . . . . . . . 198
6.2.4 Conversion from HSI to RGB . . . . . . . 198
6.3 Pseudo-Color Enhancement . . . . . . . . . . . . . 200
6.3.1 Intensity slicing . . . . . . . . . . . . . . . 201
6.3.2 Conversion from gray scale to color . . . . 202
6.3.3 Frequency domain filtering . . . . . . . . . 203
6.4 True-Color Enhancement . . . . . . . . . . . . . . 204
6.4.1 Single-component true-color
enhancement . . . . . . . . . . . . . . . . 205
6.4.2 Full-color enhancement . . . . . . . . . . . 208
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xv

Contents xv

6.5 Key Points and References for Each Section . . . . 210

Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 211
References . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Chapter 7. Image Segmentation 219
7.1 Segmentation Definition and Method
Classification . . . . . . . . . . . . . . . . . . . . . 220
7.1.1 Image segmentation definition . . . . . . . 221
7.1.2 Image segmentation algorithm
classification . . . . . . . . . . . . . . . . . 222
7.2 Differential Edge Detection . . . . . . . . . . . . . 223
7.2.1 Principle of differential edge
detection . . . . . . . . . . . . . . . . . . . 224
7.2.2 Gradient operator . . . . . . . . . . . . . . 226
7.3 Active Contour Model . . . . . . . . . . . . . . . . 231
7.3.1 Active contour . . . . . . . . . . . . . . . . 231
7.3.2 Energy function . . . . . . . . . . . . . . . 233
7.4 Thresholding Segmentation . . . . . . . . . . . . . 238
7.4.1 Principles and steps . . . . . . . . . . . . . 238
7.4.2 Threshold selection . . . . . . . . . . . . . 240
7.5 Threshold Selection Based on Transition
Region . . . . . . . . . . . . . . . . . . . . . . . . . 243
7.5.1 Transition region and effective average
gradient . . . . . . . . . . . . . . . . . . . 244
7.5.2 Extreme points of effective average
gradient and boundary of transition
region . . . . . . . . . . . . . . . . . . . . 245
7.5.3 Threshold selection . . . . . . . . . . . . . 247
7.6 Region Growing . . . . . . . . . . . . . . . . . . . 248
7.6.1 Basic method . . . . . . . . . . . . . . . . 249
7.6.2 Problems and improvements . . . . . . . . 251
7.7 Key Points and References for Each Section . . . . 252
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 255
References . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Chapter 8. Primitive Detection 265
8.1 Interest Point Detection . . . . . . . . . . . . . . . 266
8.1.1 Corner detection by second derivative . . . 266
8.1.2 Harris interest point operator . . . . . . . 268
8.1.3 Integral corner detection . . . . . . . . . . 271
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xvi

xvi 2D Computer Vision: Principles, Algorithms and Applications

8.2 Elliptical Object Detection . . . . . . . . . . . . . 277

8.2.1 Diameter bisection . . . . . . . . . . . . . 278
8.2.2 Chord-tangent method . . . . . . . . . . . 279
8.2.3 Other parameters of the ellipse . . . . . . 280
8.3 Hough Transform . . . . . . . . . . . . . . . . . . . 282
8.3.1 Point–line duality . . . . . . . . . . . . . . 283
8.3.2 Calculation steps . . . . . . . . . . . . . . 284
8.3.3 Polar coordinate equation . . . . . . . . . 287
8.4 Generalized Hough Transform . . . . . . . . . . . . 290
8.4.1 Principle of generation . . . . . . . . . . . 290
8.4.2 Complete generalized Hough
transform . . . . . . . . . . . . . . . . . . 294
8.5 Key Points and References for Each
Section . . . . . . . . . . . . . . . . . . . . . . . . 297
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 298
References . . . . . . . . . . . . . . . . . . . . . . . . . . 303

Chapter 9. Object Representation 305

9.1 Chain Code Representation of Contour . . . . . . 306
9.1.1 Chain code representation . . . . . . . . . 306
9.1.2 Chain code normalization . . . . . . . . . 307
9.2 Contour Signature . . . . . . . . . . . . . . . . . . 309
9.2.1 Distance-angle signature . . . . . . . . . . 310
9.2.2 Tangent angle-arc length signature . . . . 311
9.2.3 Slope density signature . . . . . . . . . . . 311
9.2.4 Distance-arc length signature . . . . . . . 312
9.3 Polygonal Approximation of Contour . . . . . . . . 313
9.3.1 Minimum perimeter polygon . . . . . . . . 313
9.3.2 Merging polygon . . . . . . . . . . . . . . 314
9.3.3 Splitting polygon . . . . . . . . . . . . . . 315
9.4 Hierarchical Representation of Objects . . . . . . . 316
9.4.1 Quad-tree representation . . . . . . . . . . 317
9.4.2 Binary tree representation . . . . . . . . . 319
9.5 Bounding Region of Objects . . . . . . . . . . . . 320
9.5.1 Feret box . . . . . . . . . . . . . . . . . . 320
9.5.2 Minimum enclosing rectangle . . . . . . . 321
9.5.3 Convex hull . . . . . . . . . . . . . . . . . 322
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xvii

Contents xvii

9.6 Skeleton Representation of the Object . . . . . . . 323

9.6.1 Skeleton and skeleton point . . . . . . . . 323
9.6.2 Skeleton algorithm . . . . . . . . . . . . . 324
9.7 Key Points and References for Each Section . . . . 326
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 329
References . . . . . . . . . . . . . . . . . . . . . . . . . . 337
Chapter 10. Object Description 339
10.1 Basic Contour Description Parameters . . . . . . . 340
10.1.1 Contour length . . . . . . . . . . . . . . . 340
10.1.2 Contour diameter . . . . . . . . . . . . . . 342
10.1.3 Slope, curvature, and corner point . . . . 343
10.2 Basic Region Description Parameters . . . . . . . . 344
10.2.1 Region area . . . . . . . . . . . . . . . . . 344
10.2.2 Centroid of region . . . . . . . . . . . . . . 345
10.2.3 Regional gray-scale characteristics . . . . . 345
10.3 Fourier Description of Contour . . . . . . . . . . . 347
10.3.1 Fourier description of contour . . . . . . . 347
10.3.2 Fourier description changes with
contour . . . . . . . . . . . . . . . . . . . . 350
10.4 Wavelet Description of Contour . . . . . . . . . . . 351
10.4.1 Wavelet transform basics . . . . . . . . . . 351
10.4.2 Wavelet contour descriptor . . . . . . . . . 352
10.5 Region Description with Invariant Moments . . . . 355
10.5.1 Central moment . . . . . . . . . . . . . . . 355
10.5.2 Region invariant moments . . . . . . . . . 357
10.5.3 Region affine invariant moments . . . . . . 359
10.6 Object Relationship Description . . . . . . . . . . 359
10.6.1 String description . . . . . . . . . . . . . . 359
10.6.2 Tree structure description . . . . . . . . . 362
10.7 Key Points and References for Each
Section . . . . . . . . . . . . . . . . . . . . . . . . 364
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 366
References . . . . . . . . . . . . . . . . . . . . . . . . . . 376
Chapter 11. Texture Description 377
11.1 Statistical Description of Texture . . . . . . . . . . 378
11.1.1 Co-occurrence matrix . . . . . . . . . . . . 378
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xviii

xviii 2D Computer Vision: Principles, Algorithms and Applications

11.1.2Texture descriptors based on

co-occurrence matrix . . . . . . . . . . . . 380
11.1.3 Energy-based texture descriptor . . . . . . 382
11.2 Structural Description of Texture . . . . . . . . . . 384
11.2.1 Basis of structure description method . . . 384
11.2.2 Texture tessellation . . . . . . . . . . . . . 387
11.2.3 Local binary pattern . . . . . . . . . . . . 388
11.3 Spectral Description of Texture . . . . . . . . . . . 390
11.3.1 Fourier spectrum . . . . . . . . . . . . . . 391
11.3.2 Bessel–Fourier spectrum . . . . . . . . . . 393
11.4 Key Points and References for Each Section . . . . 395
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 396
References . . . . . . . . . . . . . . . . . . . . . . . . . . 401
Chapter 12. Shape Description 403
12.1 Shape Compactness Descriptor . . . . . . . . . . . 404
12.1.1 Aspect ratio . . . . . . . . . . . . . . . . . 404
12.1.2 Form factor . . . . . . . . . . . . . . . . . 405
12.1.3 Eccentricity . . . . . . . . . . . . . . . . . 406
12.1.4 Sphericity . . . . . . . . . . . . . . . . . . 410
12.1.5 Circularity . . . . . . . . . . . . . . . . . . 410
12.1.6 Descriptor comparison . . . . . . . . . . . 411
12.2 Shape Complexity Descriptor . . . . . . . . . . . . 412
12.2.1 Simple descriptors of shape
complexity . . . . . . . . . . . . . . . . . . 413
12.2.2 Using the histogram analysis of the
blurred image to describe the shape
complexity . . . . . . . . . . . . . . . . . . 413
12.2.3 Saturation . . . . . . . . . . . . . . . . . . 414
12.3 Descriptor Based on Discrete Curvature . . . . . . 416
12.3.1 Curvature and geometric features . . . . . 416
12.3.2 Discrete curvature . . . . . . . . . . . . . 416
12.3.3 Calculation of discrete curvature . . . . . 417
12.3.4 Descriptor based on curvature . . . . . . . 420
12.4 Topological Descriptor . . . . . . . . . . . . . . . . 422
12.4.1 Euler number . . . . . . . . . . . . . . . . 422
12.4.2 Crossing number and connectivity
number . . . . . . . . . . . . . . . . . . . . 423
12.5 Key Points and References for Each Section . . . . 425
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xix

Contents xix

Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 426

References . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Chapter 13. Object Classification 433
13.1 Invariant Cross-Ratio . . . . . . . . . . . . . . . . 434
13.1.1 Cross-ratio . . . . . . . . . . . . . . . . . . 434
13.1.2 Invariant of non-collinear points . . . . . . 437
13.1.3 Symmetrical cross-ratio function . . . . . 439
13.1.4 Cross-ratio application examples . . . . . 440
13.2 Statistical Pattern Classification . . . . . . . . . . 442
13.2.1 Principle of pattern classification . . . . . 442
13.2.2 Minimum distance classifier . . . . . . . . 443
13.2.3 Optimum statistical classifier . . . . . . . 445
13.2.4 AdaBoost . . . . . . . . . . . . . . . . . . 453
13.3 Support Vector Machines . . . . . . . . . . . . . . 454
13.3.1 Linearly separable classes . . . . . . . . . 454
13.3.2 Linearly non-separable classes . . . . . . . 458
13.4 Key Points and References for Each Section . . . . 460
Self-Test Questions . . . . . . . . . . . . . . . . . . . . . 462
References . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Appendix A. Mathematical Morphology 467
A.1 Basic Set Definition . . . . . . . . . . . . . . . . . 468
A.2 Basic Operations of Binary Morphology . . . . . . 470
A.2.1 Binary dilation and erosion . . . . . . . . 470
A.2.2 Binary opening and closing . . . . . . . . 476
A.3 Combined Operations of Binary Morphology . . . 479
A.3.1 Hit-or-miss transform . . . . . . . . . . . . 480
A.3.2 Binary combination operation . . . . . . . 482
A.4 Practical Algorithm of Binary Morphology . . . . 488
A.4.1 Noise elimination . . . . . . . . . . . . . . 488
A.4.2 Corner detection . . . . . . . . . . . . . . 490
A.4.3 Contour extraction . . . . . . . . . . . . . 491
A.4.4 Region filling . . . . . . . . . . . . . . . . 492
A.4.5 Object detection and positioning . . . . . 493
A.4.6 Extraction of connected components . . . 493
A.4.7 Regional skeleton extraction . . . . . . . . 494
A.5 Key Points and References for Each Section . . . . 497
References . . . . . . . . . . . . . . . . . . . . . . . . . . 498
February 21, 2022 12:54 2D Computer Vision 9in x 6in b4515-fm page xx

xx 2D Computer Vision: Principles, Algorithms and Applications

Appendix B. Visual Constancy 501

B.1 Visual Constancy Theory . . . . . . . . . . . . . . 501
B.1.1 Various constancy . . . . . . . . . . . . . . 501
B.1.2 Retinex theory . . . . . . . . . . . . . . . 504
B.2 Application to Image Enhancement . . . . . . . . 507
B.2.1 Foggy day image enhancement . . . . . . . 507
B.2.2 Infrared image enhancement . . . . . . . . 509
B.3 Key Points and References for Each Section . . . . 511
References . . . . . . . . . . . . . . . . . . . . . . . . . . 512

Answers to Self-Test Questions 513

Index 527
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 1

Chapter 1

Computer Vision Fundamentals

Computer vision is an information discipline that uses computers

to realize human visual functions. This book mainly introduces the
basic content of computer vision, and hence can be used as an intro-
ductory textbook for learning computer vision. Some further in-depth
content can be found in “3D Computer Vision: Principle, Algorithm,
and Applications”.
The human visual process can be regarded as a complex pro-
cess from sensation (feeling the image obtained by 2D projection of
the 3D world) to perception (cognition of the content and mean-
ing of the 3D world from the 2D image). Computer vision refers to
the realization of human visual functions with computers, hoping to
make meaningful interpretations and judgments of actual targets and
scenes based on the perceived images.
This chapter gives a general introduction to computer vision to lay
some foundation for the study of the subsequent chapters. The sec-
tions of this chapter are arranged as follows:
Section 1.1 gives an overview of the basic concepts of human
vision, the difference between visual sensation and visual percep-
tion, and the steps and processes of vision.
Section 1.2 summarizes the image foundation closely related to
vision, including the representation of images and pixels, image
storage and format, image display, and printing methods (includ-
ing half-toning and dithering technology).
Section 1.3 discusses the relationship between computer vision
system and image technology, the process of the vision system and

1
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 2

2 2D Computer Vision: Principles, Algorithms and Applications

the hierarchical classification of related technologies, and intro-

duces the connection among the three layers of image engineering
technology and further classification of technical directions.
Section 1.4 introduces the structure of this book and provides a
survey for each chapter. It also briefly describes the prerequisite
knowledge and the suggestions for using this book.

1.1 Vision Basis

Computer vision is based on human vision and is closely related to

human vision. Let us introduce and discuss human vision first.

1.1.1 Vision
Human vision is generally referred to as vision for short. Vision is the
ability of humans to observe the world around them with their eyes
and to perceive the world around them with their brains. The human
visual system provides an important functional means to observe and
recognize the world, and it is also the main source of information for
humans from the outside world. According to statistics, about 75%
of the information that humans obtain from the outside world comes
from the visual system, which not only shows the huge amount of
visual information but also shows that humans have a high utilization
rate of visual information.
The following briefly introduces a few commonly used terms in
vision:

Eye (human eye): A visual organ that responds to visible light.

It mainly includes lens (eyeball), pupil, retina, and so on. The
human eye is a photosensitive organ that receives incident light. In
the eye-camera analogy, the lens, pupil, and retina of the human
eye are usually corresponding to the lens, aperture, and imaging
surface of the camera.
Retina: The last film on the peripheral wall of the human eye. It
is the light-sensitive surface layer behind the eyes, containing pho-
toreceptors and neural tissue networks. Photoreceptor cells are
distributed on the retina, which convert incident light into nerve
impulses and send them to the brain. The center of the retina
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 3

Computer Vision Fundamentals 3

is also called the fovea, where the photoreceptor cells are most
concentrated, and which is the most sensitive region in the eye.
Brain (human brain): The functional unit of the human visual
system that processes information. The brain uses nerve signals
obtained from sensors on the retina and transmitted to the brain
via the optic nerve to generate neural function patterns, which
are perceived as images.
Visible light: Electromagnetic waves within a certain wave-
length range that human eyes can feel. For normal people, the
maximum range is between 380 and 780 nm, and the minimum
range is between 400 and 700 nm. The colors felt are generally
between bright blue and white to dark red.
Color: The human visual system has different perceptions of
electromagnetic waves of different frequencies or different wave-
lengths. Color is a physical phenomenon as well as a psychological
phenomenon.
Visual acuity: The spatial resolution ability of the visual organs
(eyes). It is also the ability to finely distinguish the size and shape
of objects. Sight is usually in units of the reciprocal of the dis-
tinguishable viewing angle (1/degree). The minimum discernible
visual threshold of a normal person is about 0.5 second.
Field of view: The range of space that a person can see when
his head and eyeballs are fixed, while watching the object directly
in front of him. The maximum visual field of a normal person
is about 200 degrees (width) × 135 degrees (height).

1.1.2 Visual sensation and visual perception

From a semantic point of view, it can be considered that “vision”
includes two parts, “sensation” and “perception”, so vision can be
further divided into visual sensation and visual perception.
Human visual feeling mainly occurs in the process of imaging
the scene on the retina. It mainly involves physics, chemistry, etc.
It is to understand the basic properties of people’s response to light
(visible radiation) from a molecular perspective (such as brightness
and color). The main concerns in visual sensation are as follows:
(i) the physical properties of light, such as light quantum, light wave,
and spectrum; (ii) the degree of light stimulation of visual recep-
tors, involving photometry, eye structure, visual adaptation, visual
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 4

4 2D Computer Vision: Principles, Algorithms and Applications

intensity and sensitivity, and spatial-temporal characteristics of light;

(iii) the sensations produced by the visual system after light acts on
the retina, such as brightness and color tone.
Human visual perception is mainly related to how people trans-
form an object image into a neural response after receiving a visual
stimulus from the objective world, as well as the response manner
and the result obtained (such as shrinking the pupil or even closing
the eye when illuminated by strong light). It studies how to form
people’s representation of the external world space through vision
(such as the size of the scene, and the smooth and rough surface),
so it also has psychological factors. Visual perception is a group of
activities carried out in the human brain nerve center. It organizes
some scattered stimuli in the field of vision to form an entirety with
a certain shape and structure. Based on this, the objective world can
be understood (such as when observing a horse on the ground and a
person on it, the observer could make the judgment that the person
is riding a horse).
The objective matters that people perceived by using visual per-
ception have many characteristics. After their light stimulation, the
human visual system will produce different forms of response, so
visual perception can be further categorized as brightness percep-
tion, color perception, shape perception, space perception, and motor
perception.
From a cognitive perspective, humans not only need to obtain
information from the outside world but also need to process the infor-
mation to make judgments and decisions. Therefore, human visual
function can also be divided into two levels: visual sensation and
visual perception. Visual sensation is at a lower level, which mainly
receives external stimuli. Visual perception is at a higher level, which
converts external stimuli into meaningful content. Generally, visual
sensation completely receives external stimuli without distinction,
while visual perception must determine which parts of external stim-
uli should be combined into the “target” of concern, or analyze and
make judgments on the nature of external stimuli, so as to understand
the objective world.

1.1.3 Visual process

Vision is a complex process involving knowledge of geometry, optics,
chemistry, physiology, psychology, etc. For example, referring to the
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 5

Computer Vision Fundamentals 5

Source Reflected Light Neural Neural Scene

radiation light energy signal information information
Objective Refractive Neural
Retina Brain
world system channels

Optics Geometry Chemistry Physiology Psychology

Figure 1.1. Schematic diagram of visual process and steps.

visual flow and step diagram in Figure 1.1, the process from the
light source emitting radiation to the brain obtaining scene infor-
mation involves a series of steps. First of all, the light source illu-
minates the scene in the objective world, then it is reflected (and
may also be refracted and transmitted), and is entered in the human
eye following certain optical laws; the radiation energy received by
the human eye will pass through the refractive system inside the
human eye (including the lens and pupil, as well as cornea, aqueous
humor, and vitreous) according to the geometric law and is finally
imaged on the retina; the photoreceptor cells on the retina are stim-
ulated and respond, and the light energy is converted into the cor-
responding nerve signal according to the law of chemical reaction.
The visual information contained in the stimulus is transformed into
neural information; these neural signals are transmitted in the neural
channels of the human body by the laws of physiology, and the infor-
mation is sent to the brain under the processing of the brain’s visual
center, and combining with the laws of psychology. Only then can
people finally obtain the cognition and interpretation of the proper-
ties of objects (such as the size, position, brightness, color, movement,
trend, and situation) of the external scene.
The visual process is often broken down into three sequential
steps/processes: optical process, chemical process, and neural pro-
cessing process.

1. Optical process
The physical basis of the optical process is the human eye. From the
imaging point of view, the eyes can be simply compared with the
camera. The eye itself is a sphere with an average diameter of about
20 mm. There is a lens at the front of the sphere, which corresponds
to the lens of the camera; the pupil in front of the lens corresponds
to the aperture of the camera, which controls the light flux enter-
ing the human eye. There is a layer of retina on the inner wall of
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 6

6 2D Computer Vision: Principles, Algorithms and Applications

Retina
Lens
15 m

Fovea

x = 2.55 mm

100 m 17 mm

Figure 1.2. Optical process determines image size.

the sphere, which is a thin film containing photoreceptors and neu-

ral tissue network, which corresponds to the photosensitive surface
of the sensor in the camera (the film in the early camera). Out-
side light rays pass through the pupil and are focused on the retina
by the lens to form the image. The optical process basically deter-
mines the size of the image, which can be introduced with the help of
Figure 1.2.
In Figure 1.2, the light emitted from the scene first passes through
the lens and then forms an image on the retina. When the refrac-
tive power of the lens changes from minimum to maximum, the dis-
tance between the focal center of the lens and the retina can change
from about 17 mm to about 14 mm. For example, in Figure 1.2, the
observer looks, from a distance of 100 m, at a columnar object with
a height of 15 m. If x is used to represent the image size on the retina
in mm, and according to the geometric relationship in the figure,
15/100 = x/17, then x = 2.55 mm can be calculated.

2. Chemical process
Many light-receiving cells (light-sensitive cells) are distributed on the
surface of the retina, which can receive the energy of light and form
visual patterns. There are two types of light-receiving cells: cone cells
and rod cells. There are about 6 million to 7 million cone cells in each
eye. They are very sensitive to color. Cone cells can be divided into
three types, which have different spectral response curves to incident
radiation. Humans can use these cells to distinguish many details,
because each cell is connected to its own nerve terminal. Cone cell
vision is also called photopic vision, because cone cells only work
in a brighter environment. The combined action of the three types
of cone cells also makes people perceive color. The number of rod
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 7

Computer Vision Fundamentals 7

cells in each eye is much higher than that of cone cells, and there are
approximately 75 million to 150 million rod cells on the surface of
the retina. They have a large distribution region but low resolution
because several rod cells are connected to the same nerve terminal.
The rod cells only work in very dim light and are more sensitive
to low illumination. The rod cells mainly provide the overall vision
of the field of view, because there is only one type of rod cell, so
there is no color perception. For example, brightly colored objects in
daylight (as felt by cone cells) become colorless in moonlight because
only rod cells are working in moonlight. This phenomenon is also
called scotopic vision.
The density of cone cells is high in the fovea region. For the con-
venience of explanation, the fovea can be regarded as a 1.5 mm ×
1.5 mm square sensor matrix. The density of cone cells in this region
is about 150,000/mm2 , so it is approximated that the number of cone
cells in the fovea is about 337,000. The current electronic imaging
sensor can already concentrate even higher density photoreceptive
elements in its receiving array.
Both cone cells and rod cells are composed of pigment molecules,
which contain light-absorbing rhodopsin. This substance decom-
poses through chemical reactions after absorbing light. Once the
chemical reaction occurs, the molecules no longer absorb light. Con-
versely, if no further light passes through the retina, the chemical
reaction proceeds in reverse direction and the molecules can work
again (this conversion process often takes tens of minutes to com-
plete). When the luminous flux increases, the number of retinal
cells that are irradiated also increases, and the chemical reaction
that decomposes rhodopsin increases, so that the generated neuronal
signals become stronger. From this perspective, the retina can be
regarded as a chemical laboratory that converts optical images into
other forms of information through chemical reactions. The inten-
sity of the signal generated everywhere in the retina reflects the light
intensity of the corresponding position in the scene. It can be seen
that the chemical process basically determines the brightness or color
of the image.

3. Neuron processing
Neural processing is a process carried out in the nervous system in the
center of the brain. With the help of synapses, each retinal receiving
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 8

8 2D Computer Vision: Principles, Algorithms and Applications

Retina Retinal nerve

reception processing Cerebral cortex processing
1. Store reference image
Light Right eye Visual 2. Information processing Response
stimulate Left eye channel 3. Feature extraction
4. Decision
Retina Retinal nerve 5. Description
reception processing

Figure 1.3. Visual process flow diagram.

unit is connected to a neuron cell. Each neuron cell connects with

other cells through other synapses to form an optical neural network.
The optical neural network is further connected to the lateral regions
of the brain and to the striate cortex in the brain. There, the
sensory response to the light stimulus undergoes a series of processing
to finally form the appearance of the scene, thereby transforming
the sensation of light into the perceptual response to the scene. The
processing of the cerebral cortex has to complete a series of tasks,
from image storage to response and decision-making based on the
image.
The above three processes constitute the entire visual process,
and the overall flow diagram can be seen in Figure 1.3.
The visual process starts when the light source stimulates the
human eye. The light is reflected by the scene and enters the left
and right eyes when the visual receptors simultaneously act on the
retina to cause visual sensation. The nerve impulse generated by light
stimulation on the retina after neural processing is transmitted out of
the eye along the optic nerve fiber, and is transmitted to the cerebral
cortex through the visual channel for processing and finally causes
visual perception, or the brain responds to light stimulation to form
an image of the scene and provides explanation.

1.2 Vision and Image

A visual object is an image directly felt by the human eye, derived

from an image obtained from a scene in the objective world, and is
generally called a continuous image. Continuous here means that the
image is densely valued in space and brightness. Continuous images
are also called analog images, and the corresponding ones are dig-
ital images or discrete images. Discrete here means that the image
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 9

Computer Vision Fundamentals 9

is sampled intermittently in space and brightness. The processing

and manipulation of digital images with computers are often collec-
tively referred to as digital image technology. Because the objects
that computers can operate are all digital, digital image technology
is now directly called image technology, and digital images are also
directly called images.
Computer vision needs to be realized by many image technologies.
The objective world is three-dimensional (3D) in space, but most
imaging devices project the 3D world onto a two-dimensional (2D)
image plane, so the resulting image is 2D. 2D computer vision is
basically based on various 2D image techniques.
In the following, let’s first look at how to define and represent
the image and its basic unit, pixel, and introduce image storage and
image file format, as well as image display and printing methods.
It will be combined with the introduction of the half-toning tech-
nique and dithering technique used in printing to further grasp the
understanding of image space and amplitude (pixel gray level).

1.2.1 Images and digital images

An image can be defined as an entity that is obtained by observ-
ing the objective world in different forms and methods with vari-
ous observation systems, and can directly or indirectly act on the
human eye and thereby produce visual perception. For example,
the human visual system is an observation system, and the image
obtained through it is the image formed by the objective scene
in the human mind. Visual information comes from images, where
images are broad, such as photos, paintings, sketches, animations,
and videos. The image contains a lot of information, seeing is better
than hearing, a picture worth a thousand words, etc., illustrate this
fact.
An image can generally be represented by a 2D function f (x, y) or
a 2D array in a computer, where x and y represent the position
of a coordinate point in 2D space XY, and f represents the value
of a certain property F of the image at point (x, y). For example,
commonly used images are generally gray-level images; at this time,
f represents the gray-level value. When imaging with visible light,
the gray-level value corresponds to the observed brightness of the
objective scene.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 10

10 2D Computer Vision: Principles, Algorithms and Applications

The concept of image has been expanded in recent years.

Although generally speaking, images often refer to 2D images, 3D
images, stereo images, color images, multi-spectral images, and
multi-view images are also more and more common. Although
generally speaking, an image often refers to a single still image,
image sequences, moving images (such as TV and video), etc., have
also been used more and more extensively. Although the image is
usually displayed in the form of a gray dot matrix corresponding
to the amount of radiation, the gray level of the image may also
represent the distance or depth value in the objective world (depth
image), texture change (texture image), material absorption value
(computer tomography), etc. For example, with the advancement of
science and technological development in recent years, imaging has
expanded from visible light to other radiation bands, such as infrared,
microwave, and radio waves at the low-frequency end and ultravio-
let light, X-ray, γ rays, and cosmic rays at the high-frequency end.
In these cases, the brightness value or gray value of the image can
correspond to various radio-metrics.
Broadly speaking, an image can represent a spatial distribution of
radiant energy. This distribution can be a vector function of five vari-
ables, denoted as T (x, y, z, t, λ), where x, y, z are space variables, t
represents the time variable, λ is the spectral variable (wavelength),
and for the same group of variables, the function value T can also
be a vector (e.g., a color image includes three components, and a
multi-spectral image may include hundreds or thousands of compo-
nents). Since the actual image is limited in time and space, frequency
spectrum, and energy, T (x, y, z, t, λ) is a 5D finite function.

Example 1.1 Image examples

Figure 1.4 shows two typical public images (Lena and Cameraman;
they are also used as original images in many processing and analysis
examples in this book). The coordinate system shown in Figure 1.4(a)
is often used in screen display. Its origin O is at the upper left corner
of the image, the vertical axis marks the image row, and the hori-
zontal axis marks the image column. The coordinate system shown
in Figure 1.4(b) is often used in image calculations. Its origin is at
the lower left corner of the image, the horizontal axis is the X axis,
and the vertical axis is the Y axis. In these two figures, f (x, y) can
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 11

Computer Vision Fundamentals 11

O C Y

f (x, y)
f (x, y)

R O X
(a) (b)

Figure 1.4. Digital image and its display examples.

represent the entire image or the attribute value f of the image at

coordinates (x, y). 

1.2.2 Pixel and image representation

An image can be broken down into many units. Each basic unit
is called an image element. For 2D images, the image element is
called pixel for short. To acquire an image for a spatial scene, the
image resolution is proportional to the number of pixels contained.
The more pixels, the higher the resolution of the image, that is, the
details of the image would be more likely seen.

Example 1.2 Pixel examples

The image is composed of many pixels closely arranged, or a gray-
level image is a collection of brightness points, which can be seen as
long as the image is gradually enlarged. For example, select a small
piece (32 × 32) from Figure 1.4(b) to enlarge, see Figure 1.5(a), and
cover a 32 × 32 grid as shown in Figure 1.5(b) on it, then Figure 1.5(c)
is obtained. Each small grid in Figure 1.5(c) corresponds to a pixel,
and the gray level within the grid is consistent. Each small grid here
corresponds to a pixel. 
To represent an image, it is necessary to represent each pixel in
it. This can be achieved in multiple ways. The most commonly used
representation method is to take a 2D array f (x, y) to represent
an image, where x and y represent the position of the pixel, and
f represents the attribute value of the pixel. In this method, that
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 12

12 2D Computer Vision: Principles, Algorithms and Applications

(a) (b) (c)

Figure 1.5. Pixel examples.

is called raster representation, there is a one-to-one correspondence

between image pixels and array elements.
An image can also be represented as a 2D M ×N matrix F (where
each element represents a pixel, and M and N are the number of rows
and columns of the image, respectively):
⎡ ⎤
f11 f12 · · · f1N
⎢f ⎥
⎢ 21 f22 · · · f2N ⎥
⎢
F =⎢ . ⎥ (1.1)
. .. .. .. ⎥
⎣ . . . . ⎦
fM 1 fM 2 · · · fM N

The matrix representation method can be easily transformed into

a more convenient vector representation method. The representation
corresponding to Equation (1.1) is

F = [f 1 f2 · · · fN] (1.2)

where
 T
f i = f1i f2i ··· fM i i = 1, 2, . . . , N (1.3)

The matrix representation and vector representation are equiva-

lent, and they can be easily converted from each other.

1.2.3 Image storage and format

Images need to be represented by using a large amount of data in a
computer, and they need to be stored in a specific format.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 13

Computer Vision Fundamentals 13

1. Image memory
Image storage requires a lot of space. In image processing and
analysis systems, large capacity and fast image memory are essential.
In a computer, the smallest unit of measurement for image data is
bit. The storage capacity of the memory is commonly represented
by Bytes (B, 1B = 8 bits), Kilobytes (103 bytes, KB), Megabytes
(106 bytes, MB), Gigabytes (109 bytes, GB), Terabytes (1012 bytes,
TB), and so on. For example, to store a 1024 × 1024 8-bit image
requires 1 MB of memory. Digital storage used for image processing
and analysis can be divided into three categories:

(1) Fast memory used in the procedure of processing and analysis;

(2) Online storage for faster recall;
(3) Infrequently used database (archive) storage.

Example 1.3 Memory examples

Computer memory is a kind of memory that provides fast storage
functions. At present, the memory of a general microcomputer is
often several GB. Another type of memory that provides fast storage
is a special hardware card, also called frame buffer or video mem-
ory. It can often store multiple images and can be read at video
speed (25 or 30 images per second). It can also allow real-time zoom
in and zoom out of the image, as well as operations such as verti-
cal flipping and horizontal flipping. At present, the commonly used
frame buffer capacity can often reach dozens of GB. Flash mem-
ory, which has been widely used in recent years, has some similar-
ities with computer memory in terms of its working principle and
structure, but it can still retain the stored content after a power
failure.
Disks are more general online storages. The commonly used
Winchester disks can generally store several terabytes of data. In
recent years, magneto-optical (MO) memory has also been commonly
used, which can store gigabytes of data on a 51/4-inch optical chip.
One feature of online storage is that data need to be read frequently
and randomly, so sequential media such as tapes are generally not
used. For greater storage requirements, a disc tower can be used.
A disc tower can hold dozens to hundreds of optical discs, and uses
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 14

14 2D Computer Vision: Principles, Algorithms and Applications

mechanical devices to insert or extract optical discs from the optical

disc drive.
A characteristic of the database memory is that it requires a
very large capacity, but the data on it are read less frequently.
Generally, tapes and CDs are commonly used as database storage.
A 13-foot-long tape can store up to GB of data. However, the stor-
age life of magnetic tapes is relatively short, only seven years in
a well-controlled environment. The commonly used WORM (write-
once read-many) disc can store 6 GB of data on a 12-inch disc and
10 GB of data on a 14-inch disc. In addition, WORM discs can be
stored for more than 30 years under normal conditions. In applica-
tions that are mainly for reading, WORM discs can also be placed
in the disc tower. A WORM disc tower with a storage capacity of
terabytes can store millions of 1024 × 1024 8-bit images. 

2. Image file format

Image data are generally stored with various image file formats in
online storage and database storage. In addition to the image data
itself, the file generally needs to contain description information of
the image to facilitate the extraction and use of the image data.
There are mainly two formats for image data files: one is the
vector format and the other is raster format.
(1) In the vector format, the image is represented by a series of
line segments or a combination of line segments. The gray level
(or chromaticity) of the line segment can be uniform or vari-
able. Different gray levels can also be used in each part of the
line segment combination. A vector format file is generally like a
program file, which contains a series of commands and data. By
executing these commands, a pattern can be drawn based on the
data. Vector files are mainly used for manually drawn graphics
data files.
(2) The files that represent natural image data mainly use raster
format, which is consistent with people’s understanding of images
(an image is a collection of many image points), and is more
suitable for real images with complex color, shadow, or shape
changes. Its main disadvantage is the lack of a structure that
directly expresses the relationship between pixels and the limited
resolution of the image. The latter brings about two problems.
One is that if the image is enlarged to a certain extent, the
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 15

Computer Vision Fundamentals 15

block effect will appear, and the other is that if the image is
reduced and then restored to its original size, the image will
become blurred.

Different system platforms and software often use different image

file formats. The following briefly introduces five widely used
formats:
(1) Bitmap (BMP): BMP format is a standard image format in
the Windows environment, and its full name is Microsoft Device
Independent Bitmap (DIP). The BMP image file is also called
bitmap file, including three parts: (i) Bitmap file header (also called
table header); (ii) Bitmap information (often called palette); and
(iii) Bitmap array (image data). A bitmap file can only store one
image.
The length of the bitmap file header is fixed at 54 bytes, and it
provides information such as the type, size, print format, and starting
position of the bitmap array of the image file. The bitmap informa-
tion gives the length and width of the image, the number of bits per
pixel (it can be 1, 4, 8, 24, corresponding to the case of monochrome,
16 colors, 256 colors, and true colors), compression method, the hori-
zontal and vertical resolution of target device, and other information.
The bitmap array gives the value of each pixel in the original image
(e.g., every 3 bytes represent a pixel of a true color image, which
is the value of red, green, and blue), and its storage format can be
two kinds: compressed (only for 16-color and 256-color images) and
uncompressed. The bitmap array data are arranged starting from the
lower left corner of the image.
(2) Graphics Interchange Format (GIF): GIF format is a com-
mon image file format standard. It is an 8-bit file format (one byte
per pixel), so it can only store images up to 256 colors, and does
not support 24-bit true color images. The image data in the GIF
file is compressed. The compression algorithm used is the improved
LZW algorithm. The compression ratio provided is usually between
1:1 and 3:1. The effect is not good when there is random noise in the
image.
The GIF file structure is more complex, generally including seven
data units: file header, general palette, image data region, and four
supplementary regions (if the user just uses the GIF format to store
user image information, they may not be set). The header and image
data regions are indispensable units.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 16

16 2D Computer Vision: Principles, Algorithms and Applications

A GIF file can store multiple images (this feature is very benefi-
cial for the realization of animation on the web), so the file header
will contain global data applicable to all images and local data only
belonging to the subsequent image. When there is only one image
in the file, the global data are consistent with the local data. When
multiple images are stored, each image is collected into one image
data block, and the first byte of each block is an identifier, which
indicates the type of data block (it can be an image block, extended
block, or end symbol of the file).
(3) Joint Photographic Expert Group (JPEG): JPEG for-
mat is an international compression standard for still gray-level or
color images, especially suitable for natural photos taken, so it has
been widely used in digital cameras. The JPEG format uses a lossy
encoding format (it also has a lossless encoding format, but it has
rarely been used), and the space it can save is generally quite large.
The JPEG image file format is more complicated than other image
file formats in terms of its content and encoding method, but it does
not need to use the detailed information of each data region when
using it.
The JPEG standard itself only defines a standardized coded data
stream, and does not specify the format of image data files. Cube
Microsystems company has defined a JPEG file interchange for-
mat (JFIF). JFIF image is a JPEG image that uses either gray-level
representation or color representation with Y , Cb , and Cr compo-
nents. It contains a file header compatible with JPEG. A JFIF file
usually contains a single image. The image can be gray level and the
data part in it is a single component; it can also be color, and the
data part in it includes three components.
(4) Portable Network Graphics (PNG): PNG format is a
bitmap format with lossless compression. It uses a lossless data com-
pression algorithm derived from LZ77, and is generally used in JAVA
programs, S60 programs, or web pages. The main reason is its high
compression ratio and small file size. The PNG format has three
forms: 8-bit, 24-bit, and 32-bit. The 8-bit PNG supports two dif-
ferent types of transparency (index transparency and alpha trans-
parency). The 24-bit PNG does not support transparency (contains
only three color components R, G, and B). The 32-bit PNG adds an
8-bit transparency channel, or Alpha channel, on the basis of 24-bit,
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 17

Computer Vision Fundamentals 17

so it can show 256 levels of transparency. It this case, one can specify
the Alpha value of each pixel. When the Alpha value is 0, the pixel is
completely transparent, and when the Alpha value is 255, the pixel is
completely opaque. This allows the edges of the color image to blend
smoothly with any background, thus completely eliminating jagged
edges. This function is not available in GIF and JPEG.
Compared with the JPEG format, in case of saving text, lines,
or similar images with clear edges and large areas of the same color,
the compression effect of the PNG format would be much better than
that of the JPEG, and the image that appears damaged in the high-
contrast region, like JPEG, will not appear. However, JPEG adopts
a specific encoding method for photo images, which is suitable for the
characteristics of low-contrast, smooth color transition, more noise,
and irregular structure. Therefore, if one uses PNG instead of JPEG
at this time, the file size will increase a lot, while the improvement
of image quality is limited.
(5) Tagged Image Format File (TIFF): TIFF format is a for-
mat independent of operating system and file system (e.g., it can
be used on both the Windows environment and Macintosh environ-
ment), which is very convenient for image data exchange between
software. A TIFF image file includes file header (table header), file
directory (identification information region), and file directory entry
(image data region). There is only one file header, and at the front
end of the file. The file header gives the data storage order and the
byte offset information of the file directory. The file directory gives
information about the number of file directory entries, and has a set
of identification information that gives the address of the image data
region. The file directory is the basic unit for storing information,
also called a domain. The domains are mainly divided into five cat-
egories: basic domain, information description domain, fax domain,
document storage, and retrieval domain, as well as other domains
that are no longer recommended.
The TIFF format has a strong descriptive ability and can
develop personal identification information. The TIFF format sup-
ports images of any size, and files can be divided into five categories:
binary images, gray-level images, palette color images, full-color RGB
images, and YCb Cr images. A TIFF file can store multiple images
and multiple palette data.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 18

18 2D Computer Vision: Principles, Algorithms and Applications

1.2.4 Image display and printing

Image display refers to the display of image data in the form of
graphs (in general, the spatial arrangement of the brightness mode,
i.e., the brightness corresponding to f , is displayed in the space
(x, y)). This is also an important content of computer graphics. For
image processing, the results of the processing are mainly used for dis-
play to people. For image analysis, the analyzed data results can also
be converted into images for intuitive display with the help of com-
puter graphics technology. Therefore, image display is very important
for image processing and analysis systems.

Example 1.4 Display device examples

There are many kinds of devices that can display images. The main
display device used in common image processing and analysis sys-
tems is a TV monitor. The images input to the monitor can also be
transferred to slides, photos, or transparencies via hard copy.
In addition to TV monitors, random access cathode ray tubes
(CRTs) and various printing devices can also be used for image out-
put and display.
In CRT, the horizontal and vertical positions of the electron gun
beam can be controlled by a computer. At each deflection position,
the intensity of the electron gun beam is modulated by voltage. The
voltage of each point is proportional to the gray value corresponding
to that point. In this way, the gray-level image is transformed into a
pattern of spatial variation of light intensity, which is recorded and
displayed on the screen of the CRT.
A printing device can also be regarded as a device that displays
images, and it is generally used to output lower-resolution images. In
the early days, an easy way to print gray-level images on paper was
to take advantage of the repeatability of standard line printers. The
gray value of any point on the output image can be controlled by the
number and density of symbols or characters printed at that point.
Various thermal, inkjet, and laser printers used in recent years have
higher capabilities and can print images with higher resolution. 

1. Half-toning
The original gray level of an image often has tens to hundreds or
even thousands of levels, but some image output devices have only
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 19

Computer Vision Fundamentals 19

two gray levels, such as laser printers (either print and output black,
or do not print and output white). In order to output multi-level
gray-level images on these devices and maintain their original gray
levels, a technique called half-toning is often used. The principle of
half-toning is to use the integrated characteristics of the human eye
by printing a black dot whose size is inversely proportional to the
gray level of the pixel at each pixel position; that is, the dots printed
in the bright image region are small, while the dots printed in the
dark image region are large (the methods given in Example 1.5 have
the same principles, but the implementation techniques are differ-
ent). When the dot is small enough and the observation distance is
far enough, it is not easy for the human eye to separate the small
dots, and a relatively continuous and smooth gray-level image can
be obtained. Generally, the resolution of pictures in newspapers is
about 100 dots per inch (DPI), while the resolution of pictures in
books or magazines is about 300 dots per inch.

Example 1.5 Half-toning method

A specific implementation technique for half-toning is to first subdi-
vide the image region, and combine adjacent units to form the output
region. In this way, each output region contains several units. As long
as some units output black and others output white, different gray-
level effects can be obtained. Generally, if a unit is black at a certain
gray level, let it remain black in all outputs greater than this gray
level (here, the greater the gray level, the darker the brightness). For
example, dividing a region into 2 × 2 units can output five different
gray levels according to the manner shown in Figure 1.6, and dividing
a region into 3 × 3 units can output 10 different gray levels according
to the manner shown in Figure 1.7.
Extending this manner, to output 256 gray levels, a region needs
to be divided into 16 × 16 units. It should be noted that this manner

0 1 2 3 4

Figure 1.6. Dividing a region into 2 × 2 units to output five different gray levels.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 20

20 2D Computer Vision: Principles, Algorithms and Applications

0 1 2 3 4

5 6 7 8 9

Figure 1.7. Dividing a region into 3 × 3 units to output 10 different gray levels.

increases the amplitude resolution of the image by reducing the spa-

tial resolution of the image, so it may cause the image to be sampled
too coarsely and affect the display quality of the image. 

2. Dithering technology
The half-toning method increases the amplitude resolution of the
image by reducing the spatial resolution of the image, so the number
of gray-level outputs is subject to certain restrictions. Displaying an
image with a smaller number of gray levels may produce false con-
tours and cause image quality degradation. At this time, dither-
ing technology can be used to improve the display quality of the
coarsely quantized image by changing the amplitude value of the
image. The realization of dithering is generally to add a random
small noise d(x, y) to the original image f (x, y), that is, to add these
two values for image display. Because the values of d(x, y) and f (x, y)
have a regular relationship, it can help eliminate false contours in the
image caused by insufficient quantization (see also Example 2.11).
For example, if b is the number of bits displayed in the image (gener-
ally, b < 5), first obtain the value of d(x, y) with a uniform probability
from the following five numbers: −2(6−b) , −2(5−b) , 0, 2(5−b) , 2(6−b) ;
then, add the b most significant bits of d(x, y), which is equivalent
to adding small random noise to f (x, y), and finally the above sum
is displayed as the value of the pixel.

Example 1.6 Dithering examples

Figure 1.8 shows a set of dithering examples. Figure 1.8(a) is an
original image with 256 gray levels. Figure 1.8(b) shows the use of the
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 21

Computer Vision Fundamentals 21

(a) (b) (c) (d)

Figure 1.8. Dithering example diagram.

half-toning technique in Example 1.5 with the output image obtained

by using the units shown in Figure 1.7. Because there are only 10 gray
levels now, there are obvious false contour phenomena in the regions
where the gray-level transition is relatively slow, such as the face
and shoulders (the original smooth surface has a step change now).
Figure 1.8(c) is the result of using dithering technology to improve;
the superimposed dithering values are −2, −1, 0, 1, 2, respectively.
Figure 1.8(d) is also improved using dithering technology, but the
superimposed dithering values are −4, −2, 0, 2, 4, respectively.
It can be seen from the above examples that the use of dithering
technology can eliminate some false contours caused by using too
few gray levels. The larger the superimposed dither value, the more
obvious the effect. However, the superimposition of the dithering
value also brings noise to the image, and the larger the dithering
value, the greater the noise impact. 

1.3 Vision Systems and Image Techniques

The vision system is a system that obtains images by observing the

world and then realizes the vision function. The human visual system
includes organs such as eyes, neural networks, and cerebral cortex.
With the advancement of science and technology, there are more
and more artificial (man-made) vision systems composed of comput-
ers and electronic devices. They are trying to realize and improve
the human vision system. Artificial vision systems mainly use digital
images as input to the system.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 22

22 2D Computer Vision: Principles, Algorithms and Applications

Input Output
Objective Image Image Object Object Scenery
scene acquisition processing extraction analysis information

Computer vision system

Figure 1.9. 2D vision system flowchart.

1.3.1 Vision system flowchart

From the perspective of function completion, the 2D vision system
needs to be able to collect images of objective scenes, process the
images (preprocess), improve image quality, and then extract the
image objects corresponding to the scene of interest, as well as obtain
useful information about the objective scene through the analysis of
objects, as shown in Figure 1.9.

1.3.2 Three layers of image engineering

In order to complete the function of the vision system, a series of
techniques are required. After years of development, computer vision
technology has made great progress and there are many types of tech-
niques. There are some classification methods for these techniques,
but it seems that they are not stable and uniform. For example,
different researchers divide computer vision technology into three
layers, but they are not consistent. Someone divides it into low-level
vision, middle-level vision, and 3D vision; someone divides it into
early vision (which is further divided into two parts: only one image;
multiple images), middle-level vision, and high-level vision (geomet-
ric method); and someone else divides it into early vision (which is
divided into two portions: one image and multiple images), middle
vision, and high-level vision (which is further divided into two parts:
geometric methods; probability and inference methods).
Relatively speaking, a layered classification method for image
technology has been more consistent in the past 26 years. This
method combines various image technologies under the subject of
image engineering (a new interdisciplinary subject that systemat-
ically studies various image theories, techniques, and applications).
Image engineering can be divided into three layers: image processing,
image analysis, and image understanding. As shown in Figure 1.10,
each layer has its own characteristics.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 23

Computer Vision Fundamentals 23

Higher Image Understanding Smaller

High Symbol
(Abstraction Level)

(Data Volume)
(Semantic)

(Operand)
Middle Image Analysis Object

Low Image Processing Pixel

Lower Bigger

Figure 1.10. Three layers of image engineering.

Image processing (IP) emphasizes the transformation between

images. Although people often use image processing for referring to
a wider range of image techniques, the more narrowly defined image
processing mainly denotes various operations of images to improve
the visual effect of the image and lays the foundation for automatic
recognition, or compresses and encodes the image to reduce the nec-
essary storage space and/or transmission time to meet the require-
ments of a given communication path.
Image analysis (IA) is focused on the detection and measure-
ment of the objects of interest in the image to obtain their objective
information for establishing a description of the image. If image pro-
cessing is a process from image to image, then image analysis is
a process from image to data. The data here can be the result of
measuring the object feature, or the symbolic representation based
on the measurement. They describe mainly the characteristics and
properties of the object in the image.
The emphasis of image understanding (IU) is to further study
the nature of the objects in the image and their interrelationships on
the basis of image analysis, and to obtain an interpretation of the
meaning of the image content and the understanding of the original
objective scene. If image analysis is mainly based on the observer-
centered study of the objective world (mostly studying observable
things), then image understanding is, to a certain extent, centered on
the objective world, and with the help of knowledge and experience
to grasp and explain the whole real world (including things that are
not directly observed).
As mentioned earlier, image processing, image analysis, and image
understanding have their own characteristics in terms of abstrac-
tion level and data volume, and their operands and semantics are
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 24

24 2D Computer Vision: Principles, Algorithms and Applications

different. See Figure 1.10 for their interrelationships. Image process-

ing corresponds to relatively low-level operations, which primarily
treats the image at the pixel level, and the amount of processed data
is very large. Image analysis enters the middle level; segmentation
and feature extraction convert the original description of each pixel
of the image into a more concise description of the object in the
image. Image understanding corresponds largely to high-level opera-
tions. The operand is basically a concept symbol abstracted from the
description. Its operating process and methods have many similari-
ties with human thinking and reasoning. In addition, it can be seen
from Figure 1.10 that the amount of data gradually decreases as the
degree of abstraction increases. Specifically, the original image data
are gradually transformed into a more organized and more abstract
representation through a series of operations. In this process, seman-
tics are continuously introduced, operating objects are changed, and
the amount of data is compressed. On the other hand, high-level oper-
ations have a guiding role for low-level operations and can improve
the efficiency of low-level operations.
Among the three levels of image engineering, image processing,
and image analysis are the bases of image understanding. Their study
is relatively mature compared to image understanding and they have
been widely used in recent times. This book will mainly introduce
the content of these two levels. For more in-depth and advanced
content, see the book “3D Computer Vision: Principle, Algorithm,
and Applications”.

1.3.3 Categories of image technology

Among the three layers of image engineering, each layer includes sev-
eral technical categories (currently there are 16 categories in total),
as shown in Table 1.1.
In this book, the image processing technology mainly involves
image acquisition, image enhancement/restoration, and image multi-
resolution processing; the image analysis technology mainly involves
image segmentation and primitive detection, object representation,
description and measurement, as well as object feature extraction
and analysis, and object recognition. These categories are shown in
bold letters in Table 1.1.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 25

Computer Vision Fundamentals 25

Table 1.1. Current image technology categories studied in the three layers of
image engineering.
Three layers Image technology categories and names

Image Image acquisition (including various imaging methods, image

processing capturing, representation and storage, and camera calibration)
Image reconstruction (including image reconstruction from
projection and indirect imaging)
Image enhancement/image restoration (including
transformation, filtering, restoration, repair, replacement,
correction, and visual quality evaluation)
Image/video coding and compression (including algorithm
research, and implementation and improvement of related
international standards)
Image information security (including digital watermarking,
information hiding, image authentication, and forensics)
Image multi-resolution processing (including
super-resolution reconstruction, image decomposition and
interpolation, and resolution conversion)
Image analysis Image segmentation and primitive detection (including
edges, corners, control points, and points of interest)
Object representation, object description, feature
measurement (including binary image morphology analysis)
Object feature extraction and analysis (including color,
texture, shape, space, structure, motion, saliency, and
attributes)
Object detection and object recognition (including object
2D positioning, tracking, extraction, identification, and
classification)
Human body biological feature extraction and verification
(including detection, positioning, and recognition of human
body, face, and organs)
Image Image matching and fusion (including registration of sequence
understanding and stereo image, and mosaic)
Scene restoration (including 3D scene representation, modeling,
and reconstruction)
Image perception and interpretation (including semantic
description, scene model, machine learning, and cognitive
reasoning)
Content-based image/video retrieval (including corresponding
labeling and classification)
Spatial-temporal techniques (including high-dimensional motion
analysis, object 3D posture detection, spatial-temporal
tracking, behavior judgment, and behavior understanding)
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 26

26 2D Computer Vision: Principles, Algorithms and Applications

1.4 Overview of the Structure and Content

of This Book

As a basic textbook that introduces computer vision technology, how

to choose appropriate content and how to use the chapters in it are
also issues that need to be considered. They are discussed in the
following.

1.4.1 Structural framework and main content

According to the 2D vision system flowchart in Figure 1.9, this book
selects some related techniques for introduction. In image processing
technology, it mainly discusses various techniques of image acquisi-
tion and image enhancement/restoration, which basically correspond
to early vision or low-level vision. In image analysis technology,
it mainly discusses typical techniques of image segmentation and
primitive object detection, as well as some related techniques for
object representation, description, feature extraction, analysis, and
classification-related ones, which are mainly related to middle-level
vision.
The structural framework and main content of this book are
shown in Figure 1.11.
From the objective scene to the final extraction of scenery infor-
mation, there are four modules (solid frame): image acquisition,

Objective scene

Image (2) 2D image acquisition (B) Visual constancy

Acquisition

Image (3) Spatial domain enhancement, (4) Frequency domain

Processing enhancement, (5) Image restoration, (6) Color enhancement

(A)
Object (7) Image segmentation, (8) Primitive detection, Mathematical
Extraction (9) Object representation morphology

Object (10) Object description, (11) Texture description methods,

Analysis (12) Shape description methods, (13) Object classification

Scenery information

Figure 1.11. The structure and main content of this book.

February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 27

Computer Vision Fundamentals 27

image (pre-)processing, extraction of the object, and analysis of the

object. This book maps the techniques to be introduced to these four
modules (dotted boxes), and the numbers in parentheses correspond
to the numbers of chapters of the book. The mathematical morphol-
ogy introduced in Appendix A can be applied to different techniques
in different modules as a tool (as indicated by the arrows). The visual
constancy introduced in Appendix B is mainly related to the image
processing module (as shown by the arrow). The main content of
this book can be divided into four units as shown in Figure 1.11 (as
shown by the dotted line). The first unit includes Chapter 2, which
mainly introduces the preliminary image acquisition and represen-
tation techniques; the second unit includes Chapters 3 to 6, which
mainly introduces basic image processing techniques; the third unit
includes Chapters 7 to 9, which mainly introduces the conversion
techniques from image processing to image analysis; and the fourth
unit includes Chapters 10 to 13, which mainly introduces extended
image analysis techniques.

1.4.2 Overview of each chapter

The book has 13 chapters and two appendices.

Chapter 1 is for the introduction. It gives definitions of some basic

terms, lists examples of various images, summarizes the overall
situation of image technology, specifically introduces image rep-
resentation and display methods, image storage and file formats,
and makes some suggestions for the use of this book.
Chapter 2 introduces image acquisition methods, including geo-
metric imaging models and brightness imaging models, sampling,
and quantization for digitization, which are the keys to acquiring
digital images. In addition, the relationship between pixels in the
obtained image is also discussed.
Chapter 3 introduces spatial image enhancement methods. The
content involved includes arithmetic and logical operations on
images, gray-scale mapping technique, the use of histogram mod-
ification methods, and spatial filtering methods using pixel neigh-
borhoods.
Chapter 4 introduces the frequency domain image enhance-
ment method. Based on the summary of Fourier transform, it
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 28

28 2D Computer Vision: Principles, Algorithms and Applications

introduces a variety of low-pass, high-pass, band-stop, and band-

pass filters, and analyses the principle of homomorphic filter com-
bined with the brightness imaging model.
Chapter 5 introduces image restoration methods, analyses some
examples of image degradation, discusses basic unconstrained
restoration and constrained restoration techniques, presents
methods to correct geometric distortions, and outlines image
repair techniques.
Chapter 6 introduces color vision and color image enhancement
methods. On the basis of discussing the fundamental color vision,
physical-based and perceptual-based color models, some typical
methods for false color enhancement and true color enhancement
are given.
Chapter 7 introduces basic image segmentation methods. It first
discusses segmentation definition and method classification, and
then specifically introduces differential edge detection, active con-
tour model, threshold segmentation, threshold selection based on
transition region, and region growth technology.
Chapter 8 introduces primitive detection methods in images. Sev-
eral detection methods of interest points are discussed. The idea
of object detection is analyzed by taking ellipse as an example.
The Hough transform that can detect multiple forms of primi-
tives is also introduced and extended to the generalized Hough
transform.
Chapter 9 introduces some basic methods for object repre-
sentation, including the chain code representation of contours,
projection signatures, and polygonal approximation, as well as
the object’s hierarchical representation, surrounding region, and
skeleton representation.
Chapter 10 introduces the description techniques of the object.
In addition to some basic contour description parameters and
basic regional description parameters, it also discusses the Fourier
description and wavelet description of the contour. In addition,
the descriptions based on the region invariant moments and the
description for the object relationship are also provided.
Chapter 11 introduces the analysis methods for the object surface
texture, and discusses some typical techniques in the three types
of texture research methods that are based on statistical theory,
structural model and spectral function, respectively.
February 21, 2022 12:51 2D Computer Vision 9in x 6in b4515-ch01 page 29

Computer Vision Fundamentals 29

Chapter 12 introduces the analysis of object shape, and specif-

ically discusses four types of shape characteristic descrip-
tors, namely, shape compactness descriptors, shape complexity
descriptors, discrete curvature-based descriptors, and topological
structure descriptors.
Chapter 13 introduces the classification of object patterns.
First, the cross ratio is taken as an example to present feature
invariants, then several typical statistical pattern classifiers are
discussed, and finally the principles and characteristics of sup-
port vector machines are analyzed.
Appendix A introduces binary mathematical morphology. Based
on the review of basic set definitions, it introduces the essen-
tial operations of binary morphology, the combined operations
of binary morphology, and the practical algorithms of binary
morphology.
Appendix B introduces visual constancy, which is a type of
perceptual constancy. Based on the overview of the theory of
the retinal cortex, two applications of visual constancy in image
enhancement are also described.
At the end of each chapter and appendix, there is a section “Key
points and references for each section”. On the one hand, it sum-
marizes the central content of each section; and, on the other hand,
it provides several references for in-depth study. Except the appen-
dices, each chapter has a certain number of self-test questions (all
including hints and answers).

1.4.3 Prerequisites
From the perspective of learning image processing and analysis tech-
niques, three basic types of knowledge are more important.
(1) Mathematical knowledge: It is worth pointing out the linear alge-
bra, because the image can be represented as a matrix; it is useful
and efficient to use matrix expression to explain various pro-
cessing operations. In addition, the knowledge of statistics and
probability is also very helpful.
(2) Computer science knowledge: It is worth pointing out the com-
puter software and hardware technology, because the image
processing and analysis need to use a computer, through
Random documents with unrelated
content Scribd suggests to you:
Grigóry watched his operations with a smile of curiosity. Matréna
sniffed from time to time; the policeman had disappeared.
"So you are to attend to the lime to-day, Mr. and Mrs. Occupant.
There's a building going up alongside you, so the masons will give
you all you need for about five kopéks. And as for you, good man, if
you can't be moderate in your drinking, you must let it alone
altogether.... We-ell, good-by for the time being.... I'll look in on you
again."
And he vanished as swiftly as he had appeared, leaving as
mementos of his laughing eyes abashed and satisfied smiles on the
countenances of the Orlóff couple.
They remained silent for a minute, staring at each other, and as yet
unable to formulate the impression left by this unexpected invasion
of conscious energy into their dark, automatic life.
"A-aï!—" drawled Grigóry, shaking his head.—"So there's ... a
chemist! It is said that they are poisoning folks! But would a man
with a face like that occupy himself with that sort of thing? And then
again, his voice! And all the rest.... No, his manner was perfectly
frank, and immediately—'here now,—here I am!'—Lime ... is that
injurious? Citric acid ... what's that? Simply acid, and nothing more!
But the chief point is—cleanliness everywhere, in the air, and on the
floor, and in the slop-bucket.... Is it possible to poison a man by such
means? Akh, the devils! Poisoners, say they.... That hard-working
young fellow, hey? Fie! A workingman ought always to drink in
moderation, he says ... do you hear, Mótrya? So come now, pour me
out a little glass ... there's liquor on hand, isn't there?"
She very willingly poured him out half a cup of vódka from the
bottle, which she produced from some place known only to herself.
"That was really a nice fellow ... he had such a way of making one
like him,—" she said, smiling at the remembrance of the student.
—"But other fellows, the rest of them—who knows anything about
them? Perhaps, they actually are engaged...."
"But engaged for what, and again, by whom?—" exclaimed Grigóry.
"To exterminate the people.... They say there are so many poor
folks, that an order has been issued—to poison the superfluous
ones,"—Matréna communicated her information.
"Who says that?"
"Everybody says so.... The painters' cook said so, and a great many
other folks...."
"Well, they're fools! Would that be profitable? Just consider: they are
curing them! How is a body to understand that? They bury them!
And isn't that a loss? For a coffin is needed, and a grave, and other
things of that sort.... Everything is charged to the government
treasury.... Stuff and nonsense! If they wanted to make a clearing-
out and to reduce the number of people, they would have taken and
sent them off to Siberia—there's plenty of room for them all there!
Or to some uninhabited islands.... And after they had exiled them,
they would have ordered them to work there. Work and pay your
taxes ... understand? There's a clearing-out for you, and a very
profitable one, to boot.... Because an uninhabited island will yield no
revenue, if it isn't settled with people. And revenue is the first thing
to the public treasury, so it's not to its interest to destroy folks, and
to bury them at its expense.... Understand? And then, again, that
student ... he's an impudent creature, that's a fact, but he had more
to say about the riot; but kill people off,... no-o, you couldn't hire
him to do that for any amount of pennies! Couldn't you see at a
glance, that he wouldn't be capable of such a thing? His phiz wasn't
of that calibre...."
All day long they talked about the student, and about everything he
had told them. They recalled the sound of his laugh, his face,
discovered that one button was missing from his white coat, and
came near quarrelling over the question: 'on which side of the
breast?' Matréna obstinately maintained that it was on the right side,
her husband said—on the left, and twice cursed her stoutly, but
remembering in season, that his wife had not turned the bottle
bottom upward when she poured the vódka into the cup, he yielded
the point to her. Then they decided that on the morrow, they would
set to work to introduce cleanliness into their quarters, and again
inspired by a breath of something fresh, they resumed their
discussion of the student.
"Yes, what a go-ahead fellow he was, really now!"—said Grigóry
rapturously.—"He came in, just exactly as though he'd known us for
ten years.... He sniffed about everywhere, explained everything ...
and that was all! He didn't shout or make a row, although he's one
of the authorities, also, of course.... Akh, deuce take him! Do you
understand, Matréna, they're looking after us there, my dear. That's
evident at once.... They want to keep us sound, and nothing more,
nor less .... That's all nonsense about killing us off ... old wives'
tales.... 'How does your belly act?' says he.... And if they wanted to
kill us off, what the devil should he have wanted to know about the
action of my belly for? And how cleverly he explained all about those
... what's their name? those devils that crawl about in the bowels,
you know?"
"Something after the fashion of cock and bull stories," laughed
Matréna,—"I believe he only said that for the sake of frightening us,
and making folks more particular to keep clean...."
"Well, who knows, perhaps there's some truth in that for worms
breed from dampness.... Akh, you devil! What did he call those little
bugs? It isn't a cock and bull story at all, but ... why, I remember
what it is!... I've got the word on the tip of my tongue, but I don't
understand...."
And when they lay down to sleep, they were still talking about the
event of the day, with the same ingenuous enthusiasm with which
children communicate to one another their first experiences and the
impressions which have surprised them. Then they fell asleep, in the
midst of their discussion.
Early in the morning they were awakened. By their bedside stood
the fat cook of the painters, and her face, which was always red,
now, contrary to her wont, was gray and drawn.
"Why are you pampering yourselves?" she said hastily, making a
rather peculiar noise with her thick, red lips.—"We've got the cholera
in the court-yard.... The Lord has visited us!"—and she suddenly
burst out crying.
"Akh, you're ... lying, aren't you?" cried Grigóry.
"And I never carried out the slop-bucket last night," said Matréna
guiltily.
"My dear folks, I'm going to get my wages. I'm going away.... I'll go,
and go ... to the country," said the cook.
"Who's got it?" inquired Grigóry, getting out of bed.
"The accordeon-player! He's got it.... He drank water out of the
fountain last evening, do you hear, and he was seized in the night....
And it took him right in the belly, my good people, as though he'd
swallowed rat-poison...."
"The accordeon-player...." muttered Grigóry. He could not believe
that any disease could overcome the accordeon-player. Such a jolly,
dashing young fellow, and he had walked through the court like a
peacock, as usual, only last night.—"I'll go and take a look,"—Orlóff
decided, with an incredulous laugh.
Both women shrieked in affright:
"Grísha, why, it's catching!"
"What are you thinking of, my good man, where axe you going?"
Grigóry uttered a violent oath, thrust his legs into his trousers, and
dishevelled as he was, with shirt-collar unbuttoned, went toward the
door. His wife clutched him by the shoulder, from behind, he felt her
hands tremble, and suddenly flew into a rage, for some reason or
other.
"I'll hit you in the snout! Get away!"—he roared, and went out, after
striking his wife in the breast.
The court-yard was dark and deserted, and Grigóry, as he proceeded
toward the accordeon-player's door, was simultaneously conscious of
a chill of terror, and of a keen satisfaction at the fact that he, alone,
out of all the denizens of the house, was going to the sick
accordeon-player. This satisfaction was still further augmented when
he perceived that the tailors were watching him from the second-
story windows. He even began to whistle, wagging his head about
with a dashing air. But a little disenchantment awaited him at the
door of the accordeon-player's little den, in the shape of Sénka
Tchízhik.
Having opened the door half way, he had thrust his sharp nose into
the crack thus formed, and as was his wont, was taking his
observations, captivated to such a degree that he did not turn round
until Orlóff pulled his ear.
"Just see how it has racked him, Uncle Grigóry," he said in a whisper,
raising toward Orlóff his dirty little face, rendered still more peaked
than usual by the impressions he had undergone.—"And it's just as
though he had shrunk up and got disjointed with dryness—like a bad
cask ... by heaven!"
Orlóff, enveloped by the foul air, stood and listened in silence to
Tchízhik, endeavoring to peer, with, one eye, through the crack of
the door as it hung ajar.
"How would it do to give him some water to drink, Uncle Grigóry?"
suggested Tchízhik.
Orlóff glanced at the boy's face, which was excited almost to the
point of a nervous tremor, and felt something resembling a burst of
excitement within himself.
"Go along, fetch the water!" he ordered Tchízhik, and boldly flinging
the door wide open, he halted on the threshold, shrinking back a
little.
Athwart the mist in his eyes, Grigóry beheld Kislyakóff:—the
accordeon-player, dressed in his best, lay with his breast on the
table, which he was clutching tightly with his hands, and his feet, in
their lacquered boots, moved feebly over the wet floor.
"Who is it?" he asked hoarsely and apathetically, as though his voice
had faded, and lost all its color.
Grigóry recovered himself, and stepping cautiously over the floor, he
advanced to him, trying to speak bravely and even jestingly.
"I, brother, Mítry Pávloff.... But what are you up to ... did you
overwork last night, pray?"—he surveyed Kislyakóff attentively and
curiously, and did not recognize him.
The accordeon-player's face had grown peaked all over, his cheek-
bones projected in two acute angles, his eyes, deeply sunken in his
head, and surrounded by greenish spots, were frightfully immovable
and turbid. The skin on his cheeks was of the hue which is seen on
corpses in hot summer weather. It was a completely dead, horrible
face, and only the slow movement of the jaws showed that it was
still alive. Kislyakóff's motionless eyes stared long at Grigóry's face,
and their dead gaze put the latter in a fright. Feeling his ribs with his
hands, for some reason or other, Orlóff stood three paces distant
from the sick man, and felt exactly as though someone were
clutching him by the throat with a damp, cold hand,—were clutching
him and slowly strangling him. And he wanted to get away, as
speedily as possible, from this room, hitherto so bright and
comfortable, but now impregnated with a suffocating odor of
putrefaction, and with a strange chill.
"Well...." he was about to begin, preparatory to beating a retreat..
But the accordeon-player's gray face began to move in a strange
way, his lips, covered with a black efflorescence, parted, and he said
with his toneless voice:
"I ... am ... dying...."
The profound indifference, the inexplicable apathy of his three words
echoed in Orlóff's head and breast, like three dull blows. With a
senseless grimace on his countenance, he turned toward the door,
but Tchízhik came flying to meet him, all flushed and perspiring, with
a pail in his hand.
"Here it is ... from Spiridónoff's well ... they wouldn't let me have it,
the devils...."
He set the pail on the floor, rushed into a corner, reappeared, and
handing a glass to Orlóff, continued to prattle:
"They say you've got the cholera.... I say, well, what of that? You'll
have it too,... now it'll run the rounds, as it did in the suburbs...?
Whack! he gave me such a bang on the head that I yelled!"
Orlóff took the glass, dipped up water from the pail, and swallowed
it at one gulp. In his ears the dead words were ringing:
"I ... am ... dying...."
But Tchízhik hovered round him with swift darts, feeling himself
thoroughly in his proper sphere.
"Give me a drink...." said the accordeon-player, moving himself and
the table about on the floor.
Tchízhik hopped up to him, and held a glass of water to his black
lips. Grigóry, as he leaned against the wall by the door, listened, as
in a dream, to the sick man noisily drawing in the water; then he
heard Tchízhik propose that they should undress Kislyakóff, and put
him to bed, then the voice of the painters' cook rang out. Her broad
face, with an expression of terror and compassion, was gazing in
from the court-yard through a window, and she said in a snivelling
tone:
"You ought to give him lamp-black and rum: a tea-glass full—two
spoonfuls of lamp-black, and fill it with rum to the brim."
But some invisible person suggested olive-oil with the brine from
cucumbers, and aqua regia.
Orlóff suddenly became conscious that the heavy, oppressive gloom
within him was illuminated by some memory. He rubbed his brow
hard, as though endeavoring to increase the brilliancy of the light,
and all at once, he went swiftly thence, ran across the court-yard
and disappeared down the street.
"Heavens! And the shoemaker has got it too! He's run off to the
hospital,"—the cook commented upon his flight in a plaintively-shrill
voice.
Matréna, who was standing beside her, gazed with widely opened
eyes, and turning pale, she shook all over.
"You're mistaken," she said hoarsely, barely moving her white lips,
—"Grigóry won't fall ill of that accursed sickness.... He won't yield to
it...."
But the cook, howling wofully, had already disappeared somewhere,
and five minutes later a cluster of neighbors and passers-by was
muttering dully around the Petúnnikoff house. Over all faces the
same, identical sentiments flitted in turn: excitement, which was
succeeded by hopeless dejection, and something evil, which now
and then made way for active audacity. Tchízhik kept flying back and
forth between the court and the crowd, his bare feet twinkling, and
reporting the course of events in the accordeon-player's room.
The public, collected together in a dense knot, filled the dusty,
malodorous air of the street with the dull hum of their talk, and from
time to time a violent oath, launched at someone, broke forth from
their midst,—an oath as malicious as it was lacking in sense.
"Look ... that's Orlóff!"
Orlóff drove up to the gate on the box of a wagon with a white
canvas cover which was driven by a surly man all clad in white, also.
This man roared, in a dull bass voice:
"Get out of the way!"
And he drove straight at the people, who sprang aside in all
directions at his shout.
The aspect of this wagon, and the shout of its driver, rather subdued
the high-strung mood of the spectators,—all seemed to grow dark at
once, and many went swiftly away.
In the track of the wagon, the student who had visited the Orlóffs
made his appearance from somewhere or other. His cap had fallen
back on the nape of his neck, the perspiration streamed down his
forehead in large drops, he wore a long mantle, of dazzling
whiteness, and the lower part of its front was decorated with a large
round hole, with reddish edges, evidently just burned in some way.
"Well, Orlóff, where's the sick man?"—he asked loudly, casting a
sidelong glance at the public, which had assembled in a little niche
by the gate, and had greeted his appearance with great ill-will,
although they watched him not without curiosity.
Someone said, in a loud tone:
"Look at yourself ... you're just like a cook!".
Another voice, which was quieter and had a tinge of malice in it,
made promises:
"Just wait ... he'll give you a treat!"
There was a joker in the crowd, as there always is.
"He'll give you such soup that your belly will burst on the spot!"
A laugh rang out, though it was not merry, but obscured by a
timorous suspicion, it was not lively, though faces cleared somewhat.
"See, they ain't afraid of catching it themselves ... what's the
meaning of that?"—very significantly inquired a man with a strained
face and a glance filled with concentrated wrath.
And under the influence of this question, the countenances of the
public darkened again, and their murmurs became still duller....
"They're bringing him!"
"That Orlóff! Akh, the dog!"
"Isn't he afraid?"
"What's it to him? He's a drunkard..
"Carefully, carefully, Orlóff! Lift his feet higher ... so! Ready! Drive
off, Piótr!" ordered the student. "Tell the doctor I shall be there
soon. Well, sir, Mr. Orlóff, I request that you will help me to
exterminate the infection here.... By the way, you will learn how to
do it, in case of need.... Do you agree? Can you come?"
"I can," said Orlóff, casting a glance around him, and feeling a flood
of pride rising within him.
"And so can I," announced Tchízhik.
He had escorted the mournful wagon through the gate, and returned
just in the nick of time to offer his services. The student stared at
him through his glasses.
"Who are you, hey?"
"Apprentice ... to the house-painters...." explained Tchízhik.
"And are you afraid of the cholera?"
"I?" asked Sénka in surprise.—"The idea! I'm ... not afraid of
anything!"
"Re-eally? That's clever! Now, see here, my friends."—The student
seated himself on a cask which was lying on the ground, and rolling
himself to and fro on it, he began to say that it was indispensably
necessary that Orlóff and Tchízhik should give themselves a good
washing.
They formed a group, which was soon joined by Matréna, smiling
timidly. After her came the cook, wiping her wet eyes on her dirty
apron. In a short time, several persons from among the spectators
approached this group, as cautiously as cats approach sparrows. A
small, dense ring of men, about ten in number, formed around the
student, and this inspired him. Standing in the centre of these
people, and briskly gesticulating, he began something in the nature
of a lecture, which now awoke smiles on their faces, now aroused
their concentrated attention, now keen distrust and sceptical grins.
"The principal point in all diseases is—cleanliness of the body, and of
the air which you breathe, gentlemen,"—he assured his hearers.
"Oh Lord!" sighed the painted cook loudly.—"One must pray to Saint
Varvára the martyr to be delivered from sudden death...."
"Gentlemen live in the body and in the air, but still, they die too,"—
remarked one of the audience.
Orlóff stood beside his wife, and gazed at the face of the student,
pondering something deeply the while. Someone gave his shirt a
tug, from one side.
"Uncle Grigóry!"—whispered Sénka Tchízhik, raising himself on
tiptoe, his eyes sparkling, blazing like coals,—"now that Mítry
Pávlovitch is going to die, and he hasn't any relatives ... who'll get
his accordeon?"
"Let me alone, you imp!" Orlóff warded him off.
Sénka stepped aside, and stared through the window of the
accordeon-player's little room, searching for something in it with an
eager glance.
"Lime, tar,"—the student enumerated loudly.
On the evening of that restless day, when the Orlóffs sat down to
drink tea, Matréna asked her husband, with curiosity:
"Where did you go with the student a little while ago?"
Grigóry looked into her face with eyes obscured by something, and
different from usual, and, without replying, began to pour his tea
from his glass into his saucer.
About mid-day, after he had finished scrubbing the accordeon-
player's rooms, Grigóry had gone off somewhere with the sanitary
officer, had returned at three o'clock thoughtful and taciturn, had
thrown himself down on the bed, and there he had lain, face
upward, until tea-time, never uttering a single word all that time,
although his wife had made many efforts to draw him into
conversation. He even failed to swear at her for nagging him, and
this, in itself, was strange, she was not used to it, and it provoked
her.
With the instinct of a woman whose whole life is bound up in her
husband, she began to suspect that her husband had become
interested in something new, she was afraid of something, and
therefore, was the more passionately desirous of knowing what that
thing was.
"Perhaps you don't feel well, Grísha?"
Grigóry poured the last gulp of tea from his saucer into his mouth,
wiped his mustache with his hand, pushed his empty glass over to
his wife without haste, and knitting his brows, he said:
"I went with the student to the barracks ... yes...."
"To the cholera barracks?" exclaimed Matréna, and tremblingly, with
lowered voice, she asked: "are there many of them there?"
"Fifty-three persons, counting in our man....".."
"Well?"
"They're recovering by the score.... They can walk.... Yellow, thin...."
"Are they cholera-patients too? They're not, I suppose? ... They've
put some others in there, to justify themselves: as much as to say
—"look, we can cure!'"
"You're a fool!" said Grigóry with decision, and his eyes flashed
angrily.—"You're all stupid folks! Lack of education and stupidity—
that's all! You're enough to kill a man with your ignorance.... You
can't understand anything,"—he sharply moved toward him his glass
freshly filled with tea, and fell silent.
"Where did you get so much education?"—inquired Matréna
viciously, and sighed.
Her husband, paying not the slightest heed to her words, remained
silent, thoughtful and morose. The samovár, which had burned out,
drawled a squeaking melody, full of irritating tediousness, an odor of
oil-paints, carbolic acid, and stirred-up cesspools floated through the
windows from the court-yard. The semi-twilight, the screeching of
the samovár, and the smells—everything in the room became
densely merged with one another, forming around the Orlóffs a
setting which resembled a nightmare, while the dark maw of the
oven stared at the husband and wife exactly as though it felt itself
called upon to swallow them when a convenient opportunity should
present itself. The silence lasted for a long time. Husband and wife
nibbled away at their sugar, rattled their crockery, swallowed their
tea.[10] Matréna sighed, Grigóry tapped the table with his finger.
[10] By way of economizing, the peasants do not put sugar into
their tea, but nibble at it, and thus sweeten their mouths, an
inelegant and inconvenient, but highly satisfactory method of
operation.—Translator.

"You never saw such cleanliness as they have there!"—he suddenly

began, irritably.—"All the attendants, down to the very last one—
wear white. The sick people keep getting into the bath all the
time.... They give them wine ... six bottles and a half! As for the
food—the very smell of it would make you feel full-fed.... Care,
anxiety.... They treat them in a motherly way..? and all the rest of
it.... So they do. Please to understand: you live along upon the
earth, and not even one devil would take the trouble to spit on you,
much less call in now and then to inquire—what and how and, in
general,... what your life is like, that is to say, whether it suits you,
or whether it is the right sort for a man? Has he any means of
breathing or not? But when you begin to die—they not only do not
permit it, but even put themselves to expense. The barracks ... wine
... six bottles and a half! Haven't people any sense? For the barracks
and the wine cost a lot of money. Couldn't that same money be used
for improving life ... a little every year?"
His wife made no attempt to understand his remarks, it was enough
for her to feel that they were new, and thence to deduce, with
absolute accuracy, that something new concerning her was also in
progress in Grigóry's mind. Convinced of this, she wished to learn, as
promptly as possible, how all this concerned her. Fear was mingled
with this desire, and hope, and a sort of hostility toward her
husband.
"I suppose the people yonder know even more than you do,"—said
she, when he had finished, and pursed up her lips in a sceptical way.
Grigóry shrugged his shoulders, cast a furtive glance at her, and
then, after a pause, he began in a still more lofty tone:
"Whether they know or not, that's their business. But if I have to
die, without having seen any sort of life, I can reason about that.
Now see here, I'll tell you this: I don't want any more of this sort of
thing—that is to say, I won't consent to sit and wait for the cholera
to come and seize hold of me. I won't do it! Piótr Ivánovitch says:
'go ahead, and meet it half way! Fate is against you—but you can
oppose it,—who'll get the upper hand? It's war! That's all there is to
say about it....' So, what now? I'm going to enter the barracks as an
orderly—and that's the end of it! Understand? I'm going to walk
straight into its maw.—You may swallow me, but I'll make a play
with my feet!... I shall not earn any the less there ... twenty rubles a
month for wages, and they may add a gratuity besides.... I may
die?... that's so, but I should die sooner here. And again, it's a
change in my life...." and the excited Orlóff banged the table so
vehemently with his fist, that all the crockery bounced up and down
with a clatter.
Matréna, at the beginning of her husband's speech, had stared at
him with an expression of uneasiness, but by the time he had
finished, she had screwed up her eyes in a hostile manner.
"Did the student advise you to do that?" she asked staidly.
"I have wits of my own ... I can judge,"—for some reason, Grigóry
evaded a direct reply.
"Well, and did he advise you to separate from me?"—went on
Matréna.
"From you?"—Grigóry was somewhat disconcerted—he had not yet
succeeded in thinking out that matter. Of course, one can leave a
woman in lodgings, as is generally done, but there are different sorts
of women. Matréna, was one of the dangerous sort. One must keep
her directly under his eyes. Settling down on this thought, Orlóff
went on with a scowl:—"The student ... what ails you? You will live
here ... and I shall be earning wages ... ye-es...."
"Just so,"—said the woman briefly and calmly, and laughed with that
very significant and purely feminine smile, which is capable of
evoking in a man thoughts of jealousy which pierce his heart.
Orlóff, who was nervous and quick of apprehension, felt this, but,
being loath to betray himself, out of self-love, he flung at his wife
the curt remark:
"Quack and grunt—make up all your speeches...." and he pricked up
his ears, in anticipation of what she would say.
But she smiled again, with that exasperating smile, and preserved
silence.
"Well, how is it to be?" inquired Grigóry, in a lofty tone.
"How is what to be?" said Matréna, indifferently wiping the cups.
"Viper! None of your shiftiness—I'll damage you!" Orlóff boiled up.
—"Perhaps I'm going to my death."
"I'm not sending you ... don't go...." interrupted Matréna.
"You'd be glad to send me off, I know!" exclaimed Orlóff ironically.
She made no reply. Her silence enraged him, but he restrained
himself from his customary expression of the feelings which such
scenes called forth in him. He restrained himself under the influence
of a very venomous thought, as it appeared to him, which flashed
through his brain. He even gave vent to a malicious smile. "I know
you'd like to have me tumble down even to the very depths of hell.
Well, we shall see which of us comes off best ... yes! I, also, can
take such a course—akh, I've no patience with you!"
He sprang up from the table, snatched up his cap from the window-
sill, and went off, leaving his wife dissatisfied with her policy,
disconcerted by his threats, and with a growing feeling within her of
alarm for the future. As she gazed out of the window, she whispered
to herself:
"Oh Lord! Queen of Heaven! All-Holy Birth-Giver of God!"
Besieged by a throng of disquieting problems, she remained sitting,
for a long time, at the table, endeavoring to foresee what Grigóry
would do. Before her stood the cleanly-washed table appurtenances;
and on the principal wall of the neighboring house opposite her
windows, the setting sun cast a reddish spot; reflected from the
white wall, it penetrated into the room, and the edge of the glass
sugar-bowl which stood in front of Matréna glittered. She stared at
this faint reflection, with contracted brow, until her eyes ached.
Then, rising from her chair, she cleared away the dishes and lay
down on the bed. She felt disgusted.
Grigóry arrived when it was already entirely dark. From his very
footsteps on the stairs she decided that he was in good spirits. He
swore at the darkness in the room, called to his wife, approached
the bed, and sat down on it. His wife raised herself, and sat beside
him.
"Do you know I have something to tell you?"—asked Orlóff,
laughing.
"Well, what is it?"
"You are going to take a position also!"
"Where?" she asked, with trembling voice.
"In the same barracks with me!" announced Orlóff triumphantly.
She threw her arms round his neck, and clasping him tightly, kissed
him straight on the lips. He had not expected this, and thrust her
away. She was pretending ... she didn't want to be with him at all,
rogue that she was! The viper was pretending, she regarded her
husband as a fool....
"What are you delighted about?"—he asked roughly and
suspiciously, conscious of a desire to hurl her to the floor.
"Because I am!" she replied, boldly.
"Pretence! I know you!"
"You're my Eruslán the Brave!"[11]
[11] The hero of a seventeenth century Russian fairy-tale, after
the Persian tale of "Rustem."—Translator.
"Stop that, I tell you ... or look out for yourself!"
"You're my darling little Grísha!"
"Well, what's the matter with you, anyhow?"
When her caresses had tamed him a little, he asked her anxiously:
"But you're not afraid?"
"Why, we shall be together," she replied simply.
It pleased him to hear this. He said to her:
"You brave little creature!"
And, at the same time, he pinched her side so hard that she
shrieked.
*
The first day of the Orlóffs' service in the hospital coincided with a
very great influx of patients, and the two novices, accustomed, as
they were, to their slowly-moving existence, felt worried and
hampered in the midst of this seething activity which had seized
them in its grasp. Awkward, unable to comprehend orders,
overwhelmed by impressions, they immediately lost their heads, and
although they incessantly ran hither and thither, in the effort to
work, they hindered others rather than accomplished anything
themselves. Several times, Grigóry felt, with all his being, that he
merited a stem shout or a scolding for his incompetence, but, to his
great amazement, no one shouted at him.
When one of the doctors, a tall, black-mustached man, with a
hooked nose, and a huge wart over his right eyebrow, ordered
Grigóry to assist one of the patients to sit down in the bath-tub,
Grigóry gripped the sick man under the arms with so much zeal that
the man groaned and frowned.
"Don't break him to pieces, my dear fellow, he'll fit into the bath-tub
whole...." said the doctor seriously.
Orlóff was abashed; but the sick man, a long, gaunt fellow, laughed
with all his might, and said hoarsely: "He's new to it.... He doesn't
know how."
Another doctor, an old man, with a pointed gray beard, and large,
brilliant eyes, gave the Orlóffs instructions, when they reached the
barracks, how to treat the patients, what to do in this case and that,
how to handle the sick people in transferring them. In conclusion, he
asked them whether they had been to the bath the day before, and
gave them white aprons. This doctor's voice was soft, he spoke
rapidly; he took a great liking to the married pair, but half an hour
later they had forgotten all his instructions, overwhelmed with the
stormy life of the barracks. All about them flitted people in white,
orders were issued, caught on the fly by the orderlies, the sick
people rattled in their throats, moaned and groaned, water flowed
and splashed; and all these sounds floated on the air, which was so
thickly saturated with penetrating odors that tickled the nostrils
disagreeably, that it seemed as though every word of the doctors,
every sigh of the patients, stunk also, and irritated the nose....
At first, it seemed to Orlóff that utterly restless chaos reigned there,
wherein he could not possibly find his place, and that he would
choke, grow deaf, fall ill.... But a few hours passed, and Grigóry,
invaded by the breath of energy everywhere disseminated, pricked
up his ears, and became permeated with a mighty desire to adjust
himself to his business as speedily as possible, conscious that he
would feel calmer and easier if he could turn in company with the
rest.
"Corrosive sublimate!" shouted one doctor.
"More hot water in this bath-tub!" commanded a scraggy little
medical student, with red, inflamed eyelids.
"Here you ... what's your name? Orlóff ... yes! rub his feet.... There,
that's the way ... you understand.... So-o, so-o.... More lightly—you'll
take the skin off.... Oï, how tired I am...."
Another long-haired and pock-marked student gave Grigóry orders
and showed him how to work.
"They've brought another patient!" the news passed from one to
another.
"Orlóff, go and carry him in."
Grigóry displayed great zeal—all covered with perspiration, dizzy,
with dimmed eves and a heavy darkness in his head. At times, the
feeling of personal existence in him completely vanished under the
pressure of the mass of impressions which he underwent every
moment. The green spots under the clouded eyes on earth-colored
faces, bones which seemed to have been sharpened by the disease,
the sticky, malodorous skin, the strange convulsions of the hardly
living bodies—all this made his heart contract with grief, and caused
a nausea which he could, with difficulty, control.
Several times, in the corridor of the barracks, he caught a fleeting
glimpse of his wife; she had grown thin, and her face was gray and
abstracted. He even managed to ask her, with a voice which had
grown hoarse:
"Well, how goes it?"
She smiled faintly in reply, and silently disappeared.
A totally unaccustomed thought stung Grigóry: perhaps he had done
wrong in forcing his wife to come hither, to such filthy work. She
would fall ill of the infection.... And the next time he met her, he
shouted at her severely:
"See to it that you wash your hands often ... take care!"
"And what if I don't?"—she asked, teasingly, displaying her small,
white teeth.
This enraged him. A pretty place she had chosen for mirth, the fool!
And how mean they were, those women! But he did not succeed in
saying anything to her; catching his angry glance, Matréna went
rapidly away to the women's section.
And a minute later he was carrying his acquaintance the policeman
to the dead-house. The policeman rocked gently to and fro on the
stretcher, with his eyes fixed in a stare, from beneath contorted
brows, on the clear, hot sky. Grigóry gazed at him with dull terror in
his heart: The day before yesterday he had seen that policeman at
his post, and had even sworn at him as he went past—they had
some little accounts to settle between them. And now, here was this
man, so healthy and malicious, lying dead, all disfigured, drawn up
with convulsions.
Orlóff felt that this was not right,—why should a man be born into
the world at all, if he must die, in one day, of such a dirty disease?
He gazed down upon the policeman from above, and pitied him.
What would become of his children ... three in all? The dead man
had buried his wife a year ago, and had not yet succeeded in
marrying for the second time.
He even ached, somewhere inside, with this pity. But, all at once,
the clenched left hand of the corpse slowly moved and straightened
itself out. At the same moment, the left side of the distorted mouth,
which had been half open up to now, closed.
"Halt!"—shouted Orlóff hoarsely, setting the stretcher down on the
ground.—"Be quick!"—he said in a whisper to the orderly who was
carrying the corpse with him. The latter turned round, cast a glance
at the dead man, and said angrily to Orlóff:
"What are you lying for? Don't you understand that he's only putting
himself in order for the coffin? You see how it has twisted him up?
He can't be put into the coffin like that. Hey there, carry him along!"
"Yes, but he is moving...." protested Orlóff.
"Carry him along, do you hear, you queer man! Don't you
understand words? I tell you: he's putting himself in order,—well,
that means that he's moving. This ignorance of yours may lead you
into sin, if you don't look out.... Look lively there! Can a man make
such speeches about a dead body? That signifies a riot, brother ...
that's what it is! Understand? In other words, hold your tongue, and
don't utter a syllable to anyone about his moving,—they're all like
that. Otherwise, the sow will tell it to the boar-pig, and the boar will
tell it to the whole town, well, and the result will be a riot—'they're
burying people alive!' The populace will come here, and tear us in
bits. There'll be about enough of you left for a breakfast-roll.[12]
Understand? Shunt him here, on the left."
[12] A kalátch—a delicious and favorite form of bread, particularly
good in Moscow.—Translator.

Prónin's calm voice and leisurely gait had a sobering effect upon
Grigóry.
"Only don't let your spirits sink, my good fellow-you'll get used to it.
We're well off here. Victuals, treatment and all the rest—everything
is just as it should be. We shall all be corpses, my boy; it's the
commonest thing in life. And, in the meanwhile, brisk up, you know,
and only don't get scared—that's the chief thing! Do you drink
vódka?"
"Yes," replied Orlóff.
"Well then. Yonder in the ditch I have a little bottle, in case of need.
Come and let's swallow a little of it." They went to the pit, round the
corner of the barracks, took a drink, and Prónin, pouring some drops
of mint on sugar, gave it to Orlóff, with the words:
"Eat that, otherwise you'll smell of vódka. They're strict here about
vódka. For it's injurious to drink it, they say."
"And have you got used to things here?" Grigóry asked him.
"I should think so! I've been here from the start. A lot of folks have
died here since I've been here—hundreds, to speak plainly. It's an
uneasy life, but a good life here, to tell the truth. It's a pious work.
Like the ambulance-corps in time of war ... you've heard about the
ambulance-corps and the sisters of mercy? I watched them during
the Turkish campaign. I was at Adragan and Kars. Well, my boy,
they're purer than we are, we soldiers and people in general. We
fight, we have guns, bullets, bayonets; but they—they walk about
without any weapons, as though they were in a green garden. They
pick up our men, or a Turk, and carry them to the field-hospital. And
around them ... zh-zhee! ti-in! fi-it! Sometimes the poor ambulance
man gets it in the neck—tchik!... and that's the end of him!..."
After this conversation, and a good swallow of vódka, Orlóff plucked
up a little courage.
"You've put your hand to the rope, don't say it's too thick,"—he
exhorted himself, as he rubbed a sick man's legs. Someone behind
him entreated piteously, in a moaning voice:
"A dri-ink! Oï, my dear fellow!"
And someone gabbled:
"Oho-ho-ho! Hotter! Mis-mister doctor, it relieves me! Christ reward
you,—I can feel! Permit him to pour in some more boiling water!"
"Give him some wine!" shouted Doctor Váshtchenko.
Orlóff worked away, lending an attentive ear to what went on
around him, and found that, as a matter of fact, everything was not
so nasty and strange as it had seemed to him a little while before,
and that chaos did not reign, but a great and intelligent power was
acting regularly. But he shuddered, nevertheless, when he recalled
the policeman, and cast a furtive glance through the window of the
barracks into the yard. He believed that the policeman was dead, but
still there was an element of wavering in this belief. Wouldn't the
man suddenly spring up and shout? And he remembered that he
seemed to have heard someone tell: that one day, somewhere or
other, people who had died of the cholera leaped out of their coffins
and ran away.
As Orlóff ran to and fro in the barracks, now rubbing one patient,
now placing another in the bath-tub, he felt exactly as though gruel
were boiling in his brain. He recalled his wife: how was she getting
on yonder? Sometimes with this recollection mingled a transitory
desire to steal a minute to have a look at Matréna. But after this,
Orlóff felt, somehow, disconcerted at his desire, and exclaimed to
himself:
"Come, bustle about, you fatmeated woman! You'll dry up, never
fear.... You'll get rid of your intentions...."
He had always suspected that his wife cherished, in her heart of
hearts, intentions very insulting to him as a husband, and now and
then, when he rose in his suspicions to a sort of objectiveness, he
even admitted that there was some foundation for these intentions.
Her life, also, was tinged with yellow, and all sorts of trash creeps
into one's head with such a life. This objectiveness was generally
converted into certainty during the period of his suspicions. Then he
would ask himself: why had he found it necessary to crawl out of his
cellar into this boiling cauldron?—and he wondered at himself. But all
these thoughts worked round and round, somewhere deep within
him, and were fenced off, as it were, from the direct line of his work
by the strained attention which he devoted to the actions of the
medical staff. Never, in any sort of labor, had he beheld men wear
themselves out, as the men did here, and he reflected, more than
once, as he surveyed the exhausted faces of the doctors and
students, that all these men really did not get paid for doing
nothing!
When relieved from duty, hardly able to stand on his feet, Orlóff
went out into the court-yard of the barracks, and lay down against
its wall, under the window of the apothecary's shop. There was a
ringing in his head, there was a pain under his shoulder-blades, and
his legs ached with the gnawing pangs of fatigue. He no longer
thought of anything, or wanted anything, he simply stretched
himself out on the sod, stared at the sky, in which hung magnificent
clouds, richly adorned with the rays of sunset, and fell into a sleep
like death.
He dreamed that he and his wife were the guests of Doctor
Váshtchenko in a huge room, with rows of Vienna chairs ranged
around the walls. On the chairs all the patients from the barracks
were sitting. The doctor and Matréna were executing the "Russian
Dance" in the middle of the hall, while he himself was playing the
accordeon and laughing heartily, because the doctor's long legs
would not bend at all, and the doctor, a very grave and pompous
man, was stalking about the hall after Matréna exactly as a heron
stalks over a marsh.
All at once the policeman made his appearance in the doorway.
"Aha!" he exclaimed saturninely and menacingly.—"Did you think,
Gríshka, that I was completely dead? You're playing the accordeon,
but you dragged me out to the dead-house! Come along with me,
now! Get up!" Seized with a fit of trembling, all bathed in
perspiration, Orlóff raised himself quickly and sat on the ground.
Opposite, was squatting Doctor Váshtchenko, who said to him
reproachfully:
"What sort of an ambulance nurse are you, my friend, if you go to
sleep on the ground, and lie down on it upon your belly, to boot,
hey? Now, you'll take cold in your bowels,—you'll take to your cot,
and the first you know, you'll die.... It's not right, my friend,—you
have a place in the barracks to sleep. Why didn't they tell you so?
Besides, you are in a perspiration, and have a chill. Come along with
me, now, I'll give you something."
"I was so tired,..." muttered Orlóff.
"So much the worse. You must take care of your-self—it is a
dangerous time, and you are a valuable man."
Orlóff followed the doctor in silence along the corridor of the
barracks, in silence drank some sort of medicine out of a wine-glass,
drank something more out of another, frowned and spat.
"Come, go and have a sleep now.... Farewell for a while!" and the
doctor began to move his long, slender feet over the floor of the
corridor.
Orlóff looked after him, and suddenly ran after him, with a broad
smile.
"I thank you humbly, doctor."
"What for?" and the doctor halted.
"For the work. Now I shall try with all my might to please you!
Because your anxiety is agreeable to me ... and ... you said I was a
valuable man ... and, altogether, I'm most si-sincerely grateful to
you!" The doctor gazed intently and in surprise at the agitated face
of his hospital orderly, and smiled also.
"You're a queer fellow! However, never mind,—you'll turn out
splendidly ... genuine. Go ahead, and do your best; it will not be for
me, but for the patients. We must wrest a man from the disease,
tear him out of its paws,—do you understand me? Well, then, go
ahead and try your best to conquer the disease. And, in the mean-
while—go and sleep!"
Orlóff was soon lying on his cot, and fell asleep with a pleasing
sensation of warmth in his bowels. He felt joyful, and was proud of
his very simple conversation with the doctor.
But he sank into slumber regretting that his wife had not heard that
conversation. He must tell her to-morrow.... That devil's pepper-pot
would not believe it, in all probability.
*
"Come and drink your tea, Grísha," his wife woke him in the
morning.
He raised his head, and looked at her. She smiled at him. She was so
calm and fresh, with her hair smoothly brushed, and clad in her
white slip.
It pleased him to see her thus, and, at the same time, he reflected
that the other men in the barracks certainly must see her in the
same light.
"What do you mean—what tea? I have my own tea;—where am I to
go?" he asked gloomily.
"Come and drink it with me,"—she proposed, gazing at him with
caressing eyes.
Grigóry turned his eyes aside and said, curtly, that he would go.
She went away, but he lay down on his cot again, and began to
think.
"What a woman! She invites me to drink tea, she's affectionate....
But she has grown thin in one day." He felt sorry for her, and wanted
to do something which would please her. Should he buy something
sweet to eat with the tea? But while he was washing himself he
rejected that idea,—why pamper the woman? Let her live as she is!
They drank tea in a bright little den with two windows, which looked
out on the plain, all flooded with the golden radiance of the morning
sun. On the grass, under the windows, the dew was still glistening,
far away on the horizon in the nebulous rose-colored morning mist
stood the trees along the highway. The sky was clear, and the
fragrance of damp grass and earth floated in through the windows
from the meadows.
The table stood against the wall between the windows, and at it sat
three persons: Grigóry and Matréna with the latter's companion,—a
tall, thin, elderly woman, with a pock-marked face, and kindly gray
eyes. They called her Felitzáta Egórovna; she was unmarried, the
daughter of a Collegiate Assessor, and could not drink tea made with
water from the hospital boiler, but always boiled her own samovár.
As she explained all this to Orlóff, in a cracked voice, she hospitably
suggested that he should sit by the window, and drink his fill of "the
really heavenly air," and then she disappeared somewhere.
"Well, did you get tired yesterday?" Orlóff asked his wife.
"Just frightfully tired!" replied Matréna with animation.—"I could
hardly stand on my feet, my head reeled, I couldn't understand what
was said to me, and the first I knew, I was lying at full length on the
floor, unconscious. I barely—barely held out until relief-time came....
I kept praying; 'help, oh Lord,' I thought."
"And are you scared?"
"Of the sick people?"
"The sick people are nothing."
"I'm afraid of the dead people. Do you know...." she bent over to her
husband, and whispered to him in affright:—"they move after they
are dead.... God is my witness, they do!"
"I've se-een that!"—laughed Grigóry sceptically.—"Yesterday,
Nazároff the policeman came near giving me a box on the ear after
his death. I was carrying him to the dead-house, and he gave su-uch
a flourish with his left hand ... I hardly managed to get out of the
way ... so there now!"—He was not telling the strict truth, but it
seemed to come out that way of itself, against his will.
He was greatly pleased at this tea-drinking in the bright, clean room,
with windows opening on a boundless expanse of green plain and
blue sky. And something else pleased him, also,—not exactly his
wife, nor yet himself. The result of it all was that he wished to show
his best side, to be the hero of the day which was just beginning.
"When I start in to work—even the sky will become hot, so it will!
For there is a cause for my doing so. In the first place, there are the
people here,—there aren't any more like them on earth, I can tell
you that!"
He narrated his conversation with the doctor, and as he again
exerted his fancy, unconsciously to himself—this fact still further
strengthened his mood.
"In the second place, there's the work itself. It's a great affair, my
friend, in the nature of war, for example. The cholera and people—
which is to get the better of the other? Brains are needed, and
everything must be just so. What's cholera? One must understand
that, and then—go ahead and give it what it can't endure! Doctor
Váshtchenko says to me: 'you're a valuable man in this matter,
Orlóff,' says he. 'Don't get scared,' says he; 'and drive it up from the
patient's legs into his belly, and there,' says he, 'I'll nip it with
something sour. That's the end of it, and the man lives, and ought to
be eternally grateful to you and me, because who was it that took
him away from death? We!'"—And Orlóff proudly inflated his chest as
he gazed at his wife with kindling eyes.
She smiled pensively into his face; he was handsome, and bore a
great resemblance to his old self, the Grísha whom she had seen
some time, long ago, before their marriage.
"All of them in our division are just such hard-working, kind folks.
The woman doctor, a fa-at woman with spectacles, and then the
female medical students. They're nice people, they talk to a body so
simply, and you can understand everything they say."
"So that signifies that you're all right, satisfied?"—asked Grigóry,
whose excitement had somewhat cooled off.
"Do you mean me? Oh Lord! Judge for yourself: I get twelve rubles,
and you get twenty ... that makes thirty-eight a month![13] We're
lodged and fed! That means, that if people keep on getting sick until
the winter, how much shall we amass?... And then, God willing, we'll
raise ourselves out of that cellar...."
[13] A little less than half that amount in dollars.—Translator.
"We-ell now, that's a serious subject...." said Orlóff thoughtfully and,
after a pause, he exclaimed with the pathos of hope, as he slapped
his wife on the shoulder:—"Ekh, Matrénka,[14] isn't the sun shining
on us? Don't get scared now!"
[14] Another diminutive of Matréna.—Translator.
She flushed all over.
"If you'd only stop drinking...."
"As to that—hold your tongue! Suit your awl to your leather, your
phiz to your life.... With a different life, my conduct will be different."
"Oh Lord, if that might only happen!"—sighed his wife profoundly.
"Well now, hush up!"
"Grishenka!"[15]
[15] A third variation (Grísha, Gríshka), of Grigóry, in the
diminutive. '—Translator.
They parted with certain novel feelings toward each other, inspired
by hope, ready to work until their strength gave out, alert and
cheerful.
Two or three days passed, and Orlóff had already won several
flattering mentions as a sagacious, smart young fellow, and along
with this he observed that Prónin and the other orderlies in the
barracks bore themselves toward him with envy, and a desire to
make things unpleasant for him. He was on his guard, and he also
imbibed wrath against fat-faced Prónin, with whom he had been
inclined to strike up a friendship and to chat, "according to his soul."
At the same time, he was embittered by the plain desire of his
fellow-workers to do him some injury.—"Ekh, the rascals!" he
exclaimed to himself, and quietly gritted his teeth, endeavoring not
to let slip some convenient opportunity to pay his friends off "with as
good as they gave." And, involuntarily, his thought halted at his wife:
—with her he could talk about everything, she would not be envious
of his successes, and would not burn his boots with carbolic acid, as
Prónin had done.
All the working-days were as stormy and seething with activity as
the first had been, but Grigóry no longer became so fatigued, for he
expended his strength with more discernment with every day that
passed. He learned to distinguish the smell of the medicaments,
and, picking out from among them the odor of sulphate of ether, he
inhaled it with delight on the sly, when opportunity offered, finding
that the inhalation of ether had almost as agreeable an action as a
good glass of vódka. Catching the meaning of the medical staff at
half a word, always amiable and talkative, understanding how to
entertain the patients, he became more and more of a favorite with
the doctors and the medical students, and thus, under the combined
influence of all the impressions of his new mode of existence, a
strange, exalted mood was formed within him. He felt himself to be
a man of special qualities. In him beat the desire to do something
which should attract to him the attention of everyone, should
astonish everyone, and force them to the conviction that he had a
right to the ambition which had elevated him to such a pitch in his
own eyes. This was the singular ambition of the man who had
suddenly realized that he was a man, and who, as it were, still not
quite firmly assured of the fact, wished to confirm it, in some way, to
himself and to others; this was ambition, gradually transformed into
a thirst for some disinterested exploit.
As the result of this incentive, Orlóff performed various risky feats,
such as straining himself by carrying a heavily-built patient from his
cot to the bath-tub single-handed, without waiting for assistance
from his fellow-orderlies, nursing the very dirtiest of the patients,
behaving in a daring sort of way in regard to the possibility of
contagion, and handling the dead with a simplicity which sometimes
passed over into cynicism. But all this did not satisfy him; he longed
for something on a greater scale, and this longing burned
incessantly within him, tortured him, and, at last, drove him to
anguish.
Then he poured out his soul to his wife, because he had no one else.
One evening, when he and his wife were relieved from duty, they
went out into the fields, after they had drunk tea. The barracks
stood far away from the town, in the middle of a long, green plain,
bounded on one side by a dark strip of forest, on the other by the
line of buildings in the town; on the north the plain extended into
the far distance, and there its verdure became merged with the dull-
blue horizon; on the south it was intersected by a precipitous
descent to the river, and along the verge of this precipice ran the
highway, along which, at equal distances one from another, stood
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com