尼克松和阿瓜多编著的《计算机视觉特征提取与图像处理》是由英国南安普顿大学的Mark Nixon教授和Sportradar公司的Alberto S.Aguado在第二版的基础上，于2012年9月推出的最新改版之作（第3版）。本次改版，主要的变化是将高级特征提取，分为固定形状匹配与可变形形状分析两部分，并增加了新的一章内容：运动对象检测与描述。具体地，在简要介绍计算机视觉的基础概念和基本的图像处理运算后，重点讨论了低级和高级的特征提取，包括边缘检测、固定形状匹配和可变形形状分析。

尼克松和阿瓜多编著的《计算机视觉特征提取与图像处理》是由英国南安普顿大学的MarkNixon教授和Sportradar公司的Alberto S.Aguado在第二版的基础上，于2012年9月推出的最新改版之作。本次改版，主要的变化是将高级特征提取，分为固定形状匹配与可变形形状分析两部分，并增加了新的一章内容：运动对象检测与描述。具体地，在简要介绍计算机视觉的基础概念和基本的图像处理运算后，重点讨论了低级和高级的特征提取，包括边缘检测、固定形状匹配和可变形形状分析。此外，对目标描述，纹理描述、分割及分类，以及运动对象检测等都进行深入的阐述。它突出了计算机视觉的主要问题——特征提取，以清晰、简洁的语言，阐述了图像处理和计算机视觉的基础理论与技术。

《计算机视觉特征提取与图像处理》可作为高等学校电子工程、计算机科学、计算机工程等专业本科生的双语教材，也可以作为图像、视频信号处理，模式识别和计算机视觉研究方向的博士生、硕士研究生，以及相关专业的科研工作者的参考用书。

Preface

About the authors

CHAPTER 1 Introduction

 1.1 Overview

 1.2 Human and computer vision

 1.3 The human vision system

 1.3.1 The eye

 1.3.2 The neural system

 1.3.3 Processing

 1.4 Computer vision systems

 1.4.1 Cameras

 1.4.2 Computer interfaces

 1.4.3 Processing an image

 1.5 Mathematical systems

 1.5.1 Mathematical tools

 1.5.2 Hello Matlab, hello images!

 1.5.3 Hello Mathcad!

 1.6 Associated literature

 1.6.1 Journals, magazines, and conferences

 1.6.2 Textbooks

 1.6.3 The Web

 1.7 Conclusions

 1.8 References

CHAPTER 2 Images, Sampling, and Frequency

 Domain Processing

 2.1 Overview

 2.2 Image formation

 2.3 The Fourier transform

 2.4 The sampling criterion

 2.5 The discrete Fourier transform

 2.5.1 1D transform

 2.5.2 2D transform

 2.6 Other properties of the Fourier transform

 2.6.1 Shift invariance

 2.6.2 Rotation

 2.6.3 Frequency scaling

 2.6.4 Superposition (linearity)

 v

 2.7 Transforms other than Fourier

 2.7.1 Discrete cosine transform

 2.7.2 Discrete Hartley transform

 2.7.3 Introductory wavelets

 2.7.4 Other transforms

 2.8 Applications using frequency domain properties

 2.9 Further reading

 2.10 References

CHAPTER 3 Basic Image Processing Operations 83

 3.1 Overview

 3.2 Histograms

 3.3 Point operators

 3.3.1 Basic point operations

 3.3.2 Histogram normalization

 3.3.3 Histogram equalization

 3.3.4 Thresholding

 3.4 Group operations

 3.4.1 Template convolution

 3.4.2 Averaging operator

 3.4.3 On different template size

 3.4.4 Gaussian averaging operator

 3.4.5 More on averaging

 3.5 Other statistical operators

 3.5.1 Median filter

 3.5.2 Mode filter

 3.5.3 Anisotropic diffusion

 3.5.4 Force field transform

 3.5.5 Comparison of statistical operators

 3.6 Mathematical morphology

 3.6.1 Morphological operators

 3.6.2 Gray-level morphology

 3.6.3 Gray-level erosion and dilation

 3.6.4 Minkowski operators

 3.7 Further reading

 3.8 References

CHAPTER 4 Low-Level Feature Extraction (including

 edge detection)

 4.1 Overview

 4.2 Edge detection

 4.2.1 First-order edge-detection operators

 4.2.2 Second-order edge-detection operators

 vi Contents

 4.2.3 Other edge-detection operators

 4.2.4 Comparison of edge-detection operators

 4.2.5 Further reading on edge detection

 4.3 Phase congruency

 4.4 Localized feature extraction

 4.4.1 Detecting image curvature (corner extraction)

 4.4.2 Modern approaches: region/patch analysis

 4.5 Describing image motion

 4.5.1 Area-based approach

 4.5.2 Differential approach

 4.5.3 Further reading on optical flow

 4.6 Further reading

 4.7 References

CHAPTER 5 High-Level Feature Extraction: Fixed Shape

 Matching

 5.1 Overview

 5.2 Thresholding and subtraction

 5.3 Template matching

 5.3.1 Definition

 5.3.2 Fourier transform implementation

 5.3.3 Discussion of template matching

 5.4 Feature extraction by low-level features

 5.4.1 Appearance-based approaches

 5.4.2 Distribution-based descriptors

 5.5 Hough transform

 5.5.1 Overview

 5.5.2 Lines

 5.5.3 HT for circles

 5.5.4 HT for ellipses

 5.5.5 Parameter space decomposition

 5.5.6 Generalized HT

 5.5.7 Other extensions to the HT

 5.6 Further reading

 5.7 References

CHAPTER 6 High-Level Feature Extraction: Deformable

 Shape Analysis

 6.1 Overview

 6.2 Deformable shape analysis

 6.2.1 Deformable templates

 6.2.2 Parts-based shape analysis

 Contents vii

 6.3 Active contours (snakes)

 6.3.1 Basics

 6.3.2 The Greedy algorithm for snakes

 6.3.3 Complete (Kass) snake implementation

 6.3.4 Other snake approaches

 6.3.5 Further snake developments

 6.3.6 Geometric active contours (level-set-based

 approaches)

 6.4 Shape skeletonization

 6.4.1 Distance transforms

 6.4.2 Symmetry

 6.5 Flexible shape models—active shape and active

 appearance

 6.6 Further reading

 6.7 References

CHAPTER 7 Object Description

 7.1 Overview

 7.2 Boundary descriptions

 7.2.1 Boundary and region

 7.2.2 Chain codes

 7.2.3 Fourier descriptors

 7.3 Region descriptors

 7.3.1 Basic region descriptors

 7.3.2 Moments

 7.4 Further reading

 7.5 References

CHAPTER 8 Introduction to Texture Description,

 Segmentation, and Classification

 8.1 Overview

 8.2 What is texture?

 8.3 Texture description

 8.3.1 Performance requirements

 8.3.2 Structural approaches

 8.3.3 Statistical approaches

 8.3.4 Combination approaches

 8.3.5 Local binary patterns

 8.3.6 Other approaches

 8.4 Classification

 8.4.1 Distance measures

 8.4.2 The k-nearest neighbor rule

 8.4.3 Other classification approaches

 viii Contents

 8.5 Segmentatio

 8.6 Further reading

 8.7 References2

CHAPTER 9 Moving Object Detection and Description

 9.1 Overview

 9.2 Moving object detection

 9.2.1 Basic approaches

 9.2.2 Modeling and adapting to the (static) background

 9.2.3 Background segmentation by thresholding

 9.2.4 Problems and advances

 9.3 Tracking moving features

 9.3.1 Tracking moving objects

 9.3.2 Tracking by local search

 9.3.3 Problems in tracking

 9.3.4 Approaches to tracking

 9.3.5 Meanshift and Camshift

 9.3.6 Recent approaches

 9.4 Moving feature extraction and description

 9.4.1 Moving (biological) shape analysis

 9.4.2 Detecting moving shapes by shape matching

 in image sequences

 9.4.3 Moving shape description

 9.5 Further reading

 9.6 References

CHAPTER 10 Appendix 1: Camera Geometry Fundamentals

 10.1 Image geometry

 10.2 Perspective camera

 10.3 Perspective camera model

 10.3.1 Homogeneous coordinates and projective

 geometry

 10.3.2 Perspective camera model analysis

 10.3.3 Parameters of the perspective camera model

 10.4 Affine camera

 10.4.1 Affine camera model

 10.4.2 Affine camera model and the perspective

 projection

 10.4.3 Parameters of the affine camera model

 10.5 Weak perspective model

 10.6 Example of camera models

 10.7 Discussion

 10.8 References

 Contents ix

CHAPTER 11 Appendix 2: Least Squares Analysis

 11.1 The least squares criterion

 11.2 Curve fitting by least squares1

CHAPTER 12 Appendix 3: Principal Components Analysis

 12.1 Principal components analysis

 12.2 Data

 12.3 Covariance

 12.4 Covariance matrix

 12.5 Data transformation

 12.6 Inverse transformation1

 12.7 Eigenproblem

 12.8 Solving the eigenproblem

 12.9 PCA method summary

 12.10 Example

 12.11 References

CHAPTER 13 Appendix 4: Color Images1

 13.1 Color images2

 13.2 Tristimulus theory2

 13.3 Color models

 13.3.1 The colorimetric equation

 13.3.2 Luminosity function

 13.3.3 Perception based color models: the CIE RGB

 and CIE XYZ

 13.3.4 Uniform color spaces: CIE LUV and CIE LAB

 13.3.5 Additive and subtractive color models: RGB

 and CMY

 13.3.6 Luminance and chrominance color models:

 YUV, YIQ, and YCbCr

 13.3.7 Perceptual color models: HSV and HLS

 13.3.8 More color models

 13.4 References

 Index