Digital Image Processing and Analysis: Computer Vision and Image Analysis, 4th Edition

✍ Scribed by Scott E Umbaugh

• Publisher: CRC Press
• Year: 2024
• Tongue: English
• Leaves: 441
• Edition: 4
• Category: Library

✦ Synopsis

Computer Vision and Image Analysis, focuses on techniques and methods for image analysis and their use in the development of computer vison applications. The field is advancing at an ever increasing pace, with applications ranging from medical diagnostics to space exploration. The diversity of applications is one of the driving forces that make it such an exciting field to be involved in for the 21st century. This book presents a unique engineering approach to the practice of computer vision and image analysis, which starts by presenting a global model to help gain an understanding of the overall process, followed by a breakdown and explanation of each individual topic. Topics are presented as they become necessary for understanding the practical imaging model under study, which provides the reader with the motivation to learn about and use the tools and methods being explored.

The book includes chapters on image systems and software, image analysis, edge, line and shape detection, image segmentation, feature extraction and pattern classification. Numerous examples, including over 500 color images are used to illustrate the concepts discussed. Readers can explore their own application development with any programming languages, including C/C++, MATLAB, Python, and R, and software is provided for both the Windows/C/C++ and MATLAB environments.

The primary reasons for a new edition of the book are (1) to create two separate volumes corresponding to each application area, (2) to include new material for the CVIPtools MATLAB Toolbox GUI, (3) to update the book for this rapidly changing field and (4) to add more examples and images. After many requests from CVIPtools users regarding the MATLAB version, we decided it is a good idea to expand the MATLAB version to include a graphical user interface (GUI), similar to the one in the standard Windows CVIPtools. Although there are a number of imaging tools currently available for use with MATLAB, we believe that there are important aspects of the CVIPtools libraries that are unique. The comprehensive nature of the libraries and the fact that they have been in use by the imaging community for over 25 years are important considerations which guided us in the decision to create the CVIP MATLAB Toolbox and its new GUI.

Who Will Use This Book?
Computer Vision and Image Analysis is intended for use by the academic community in teaching and research, as well as working professionals performing research and development in the commercial sectors. This includes all areas of computer vision and image analysis, and presents a computer vision system as a deployed image analysis system. It will be useful to academics and practicing engineers, consultants and programmers, as well as those in the graphics fields, medical imaging professionals, multimedia specialists and others. The book can be used for self-study and is of interest to anyone involved with developing computer vision applications, whether they are engineers, geographers, biologists, oceanographers or astronomers. At the university, it can be used as a text in standard computer vision senior-level or graduate course, or it may be used at any level in an applications-oriented course. One essential component that is missing from standard theoretical textbooks is a conceptual presentation of the material, which is fundamental to gaining a solid understanding of these complex topics. Additionally, this book provides the theory necessary to understand the foundations of computer vision, as well as that which is needed for new algorithm development.

The prerequisites for the book are an interest in the field, a basic background in computers and a basic math background provided in an undergraduate science or engineering program. Knowledge of the C family of programming languages, including C, C++ and C#, and/or MATLAB experience will be necessary for those intending to develop algorithms at the application level. The book is written so that readers without a complete engineering background can learn to use the tools and achieve a conceptual understanding of the material.

✦ Table of Contents

Cover
Half Title
Title Page
Copyright Page
Dedication
Table of Contents
Preface
Acknowledgments
Author
1 Digital Image Processing and Analysis
1.1 Introduction
1.2 Image Analysis and Computer Vision Overview
1.3 Digital Imaging Systems
1.4 Image Formation and Sensing
1.4.1 Visible Light Imaging
1.4.2 Imaging Outside the Visible Range of the EM Spectrum
1.4.3 Acoustic Imaging
1.4.4 Electron Imaging
1.4.5 Laser Imaging
1.4.6 Computer-Generated Images
1.5 Image Representation
1.5.1 Binary Images
1.5.2 Gray-Scale Images
1.5.3 Color Images
1.5.4 Multispectral and Multiband Images
1.5.5 Digital Image File Formats
1.6 Key Points
1.7 References and Further Reading
1.8 Exercises
2 Computer Vision Development Tools
2.1 Introduction and Overview
2.2 CVIPtools Windows GUI
2.2.1 Image Viewer
2.2.2 Analysis Window
2.2.3 Utilities Window
2.2.4 Help Window
2.2.5 Development Tools
2.3 CVIPlab for C/C++ Programming
2.3.1 Toolkit, Toolbox Libraries and Memory Management in C/C++
2.3.2 Image Data and File Structures
2.4 The MATLAB CVIP Toolbox
2.4.1 Help Files
2.4.2 M-Files
2.4.3 CVIPtools for MATLAB GUI
2.4.4 CVIPlab for MATLAB
2.4.5 Vectorization
2.4.6 Using CVIPlab for MATLAB
2.4.7 Adding a Function
2.4.8 A Sample Batch Processing M-File
2.4.9 VIPM File Format
2.5 References and Further Reading
2.6 Introductory Programming Exercises
2.7 Computer Vision and Image Analysis Projects
3 Image Analysis and Computer Vision
3.1 Introduction
3.1.1 Overview
3.1.2 System Model
3.2 Preprocessing
3.2.1 Region of Interest Geometry
3.2.2 Arithmetic and Logic Operations
3.2.3 Enhancement with Spatial Filters
3.2.4 Enhancement with Histogram Operations
3.2.5 Image Quantization
3.3 Binary Image Analysis
3.3.1 Thresholding Bimodal Histograms
3.3.2 Connectivity and Labeling
3.3.3 Basic Binary Object Features
3.3.4 Computer Vision: Binary Object Classification
3.4 Key Points
3.5 References and Further Reading
3.6 Exercises
3.6.1 Programming Exercises
3.7 Supplementary Exercises
3.7.1 Supplementary Programming Exercises
4 Edge, Line and Shape Detection
4.1 Introduction and Overview
4.2 Edge Detection
4.2.1 Gradient Operators
4.2.2 Compass Masks
4.2.3 Thresholds, Noise Mitigation and Edge Linking
4.2.4 Advanced Edge Detectors
4.2.5 Edges in Color Images
4.2.6 Edge Detector Performance
4.3 Line Detection
4.3.1 Hough Transform
4.3.2 Postprocessing
4.4 Corner and Shape Detection
4.4.1 Corner Detection
4.4.2 Shape Detection with the Hough Transform
4.5 Key Points
4.6 References and Further Reading
4.7 Exercises
4.7.1 Programming Exercises
4.8 Supplementary Exercises
4.8.1 Supplementary Programming Exercises
5 Segmentation
5.1 Introduction and Overview
5.1.1 Segmentation System Model and Preprocessing
5.1.2 Image Segmentation Categories
5.2 Region Growing and Shrinking
5.3 Clustering Techniques
5.4 Boundary Detection
5.5 Deep Learning Segmentation Methods
5.5.1 Convolution Neural Networks
5.6 Combined Segmentation Approaches
5.7 Morphological Filtering
5.7.1 Erosion, Dilation, Opening, Closing
5.7.2 Hit-or-Miss Transform, Thinning and Skeletonization
5.7.3 Iterative Modification
5.8 Segmentation Evaluation Methods
5.8.1 Binary Object Shape Comparison Metrics
5.8.2 Subjective Methods for Complex Images
5.8.3 Objective Methods for Complex Images
5.9 Key Points
5.10 References and Further Reading
5.11 Exercises
5.11.1 Programming Exercises
5.12 Supplementary Exercises
5.12.1 Supplementary Programming Exercises
6 Feature Extraction and Analysis
6.1 Introduction and Overview
6.1.1 Feature Extraction
6.2 Shape Features
6.3 Histogram Features
6.4 Color Features
6.5 Fourier Transform and Spectral Features
6.6 Texture Features
6.7 Region-Based Features: SIFT/SURF/GIST
6.8 Feature Extraction with CVIPtools
6.9 Feature Analysis
6.9.1 Feature Vectors and Feature Spaces
6.9.2 Distance and Similarity Measures
6.9.3 Data Preprocessing
6.10 Key Points
6.11 References and Further Reading
6.12 Exercises
6.12.1 Programming Exercises
6.13 Supplementary Exercises
6.13.1 Supplementary Programming Exercises
7 Pattern Classification
7.1 Introduction
7.2 Algorithm Development: Training and Testing Methods
7.3 Nearest Neighbor (NN), K-NN, Nearest Centroid, Template Matching
7.4 Bayesian, Support Vector Machines, Random Forest Classifiers
7.5 Neural Networks and Deep Learning
7.6 Cost/Risk Functions and Success Measures
7.7 Pattern Classification Tools: Python, R, MATLAB and CVIPtools
7.7.1 Python
7.7.2 R: Bayesian Modeling and Visualization Tools
7.7.3 MATLAB: Statistics and Machine Learning
7.7.4 CVIPtools
7.8 Key Points
7.9 References and Further Reading
7.10 Exercises
7.10.1 Programming Exercises
7.11 Supplementary Exercises
7.11.1 Supplementary Programming Exercises
8 Application Development Tools
8.1 Introduction and Overview
8.2 CVIP Algorithm Test and Analysis Tool
8.2.1 Overview and Capabilities
8.2.2 How to Use CVIP-ATAT
8.2.2.1 Running CVIP-ATAT
8.2.2.2 Creating a New Project
8.2.2.3 Inserting Images
8.2.2.4 Inputting an Algorithm
8.2.2.5 Executing an Experiment
8.3 CVIP-ATAT: Application Development Necrotic Liver Tissue
8.3.1 Introduction and Overview
8.3.2 The Algorithm
8.3.3 Conclusion
8.4 CVIP-ATAT: Application Development with Fundus Images
8.4.1 Introduction and Overview
8.4.2 The New Algorithm
8.4.3 Conclusion
8.5 CVIP-ATAT: Automatic Mask Creation of Gait Images
8.5.1 Introduction
8.5.2 Gait Analysis Images
8.5.3 Preprocessing
8.5.4 Algorithm Combinations
8.5.5 Results Analysis
8.5.6 Conclusion
References
8.6 CVIP Feature Extraction and Pattern Classification Tool
8.6.1 Overview and Capabilities
8.6.2 How to Use CVIP-FEPC
8.6.2.1 Running CVIP-FEPC
8.6.2.2 Creating a New Project
8.6.2.3 Entering Classes in CVIP-FEPC
8.6.2.4 Adding Images and Associated Classes
8.6.2.5 Applying Feature Extraction and Pattern Classification
8.6.2.6 Running a Single Test with Training and Test Sets
8.6.2.7 The Result File
8.6.2.8 Running a Leave-One-Out Test in Combinatoric Mode
8.7 CVIP-FEPC: Application Development with Thermograms
8.7.1 Introduction and Overview
8.7.2 Setting Up Experiments
8.7.3 Running the Experiments and Analyzing Results
8.7.4 Conclusion
8.8 CVIP-FEPC: Identification of Bone Cancer in Canine Thermograms
8.8.1 Introduction
8.8.2 Clinical Application Development
8.8.2.1 Image Database
8.8.2.2 Feature Extraction and Pattern Classification
8.8.2.3 Experimental Setup
8.8.3 Results and Discussion
8.8.4 Conclusion
References
8.9 MATLAB CVIP Toolbox GUI: Detection of Syrinx in Canines with Chiari Malformation via Thermograms
8.9.1 Introduction
8.9.2 Material and Methods
8.9.2.1 Image Data Acquisition
8.9.2.2 ROI Extraction
8.9.2.3 MATLAB
8.9.2.4 CVIPtools
8.9.3 MATLAB CVIP Toolbox
8.9.3.1 Feature Extraction and Pattern Classification
8.9.3.2 Features
8.9.3.3 Data Normalization Methods
8.9.3.4 Distance Metrics
8.9.3.5 Classification Methods
8.9.4 CVIPtools MATLAB Toolbox GUI
8.9.4.1 Feature Extraction Using the MATLAB GUI
8.9.4.2 Pattern Classification Using MATLAB GUI
8.9.5 Results and Discussion
8.9.6 Conclusion
References
Index