DIGITAL IMAGE PROCESSING AND ANALYSIS Human and Computer Vision Applications with CVIPtools SECOND EDITION SCOTT E UMBAUGH • Uffi\ CRC Press Taylor &. Francis Group Boca Raton London New York CRC Press is an imprint of the Taylor & Francis Group, an informa business
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DIGITAL IMAGE PROCESSING AND ANALYSIS Human and Computer Vision Applications with CVIPtools
SECOND EDITION
SCOTT E UMBAUGH
•
Uffi\ CRC Press Taylor &. Francis Group
Boca Raton London New York
CRC Press is an imprint of the Taylor & Francis Group, an informa business
Contents
Preface xv Acknowledgments xix Author xxi
Section I Introduction to Digital Image Processing and Analysis
1. Digital Image Processing and Analysis 3 1.1 Overview 3 1.2 Image Analysis and Computer Vision 5 1.3 Image Processing and Human Vision 8 1.4 Key Points 12 Exercises 13 References 13 Further Reading 14
2. Computer Imaging Systems 15 2.1 Imaging Systems Overview 15 2.2 Image Formation and Sensing 20
2.2.1 Visible Light Imaging 21 2.2.2 Imaging outside the Visible Range of the EM Spectrum 28 2.2.3 Acoustic Imaging 30 2.2.4 Electron Imaging 32 2.2.5 Laser Imaging 33 2.2.6 Computer-Generated Images 34
6. Feature Analysis and Pattern Classification 335 6.1 Introduction and Overview 335 6.2 Feature Extraction 336
6.2.1 Shape Features 337 6.2.2 Histogram Features 341 6.2.3 Color Features 347 6.2.4 Spectral Features 347 6.2.5 Texture Features 349 6.2.6 Feature Extraction with CVIPtools 354
6.3 Feature Analysis 357 6.3.1 Feature Vectors and Feature Spaces 357 6.3.2 Distance and Similarity Measures 359 6.3.3 Data Preprocessing 364
6.4 Pattern Classification 368 6.4.1 Algorithm Development: Training and Testing Methods 368 6.4.2 Classification Algorithms and Methods 370 6.4.3 Cost/Risk Functions and Success Measures 373 6.4.4 Pattern Classification with CVIPtools 376
Section IV Programming and Application Development with CVIPtools
11. CVIPlab 715 11.1 Introduction to CVIPlab 715 11.2 Toolkits, Toolboxes, and Application Libraries 721 11.3 Compiling and Linking CVIPlab 722
11.3.1 How to Build the CVIPlab Project with Microsoft's Visual C++® 2008 722
11.3.2 Mechanics of Adding a Function with Microsoft's Visual C++® 2008 724
11.3.3 Using CVIPlab in the Programming Exercises with Microsoft's Visual C++® 2008 728
11.3.4 Using Microsoft's Visual C++® 2010 731 11.4 Image Data and File Structures 734 11.5 CVIP Projects 739
11.5.1 Digital Image Analysis and Computer Vision Projects 739 11.5.2 Digital Image Processing and Human Vision Projects 741
12. Application Development 743 12.1 Introduction and Overview 743 12.2 CVIP Algorithm Test and Analysis Tool 744
12.2.1 Overview and Capabilities 744 12.2.2 How to Use CVIP-ATAT 744
12.2.2.1 Running CVIP-ATAT 744 12.2.2.2 Creating a New Project 744 12.2.2.3 Inserting Images 745 12.2.2.4 Inputting an Algorithm 747 12.2.2.5 Performing an Algorithm Test Run 751 12.2.2.6 Comparing Images 751
12.2.3 Application Development Example with Fundus Images 754 12.2.3.1 Introduction and Overview 754 12.2.3.2 New Algorithm 755 12.2.3.3 Conclusion 760
12.3 CVIP Feature Extraction and Pattern Classification Tool 761 12.3.1 Overview and Capabilities 761 12.3.2 How to Use CVIP-FEPC 761
12.3.2.1 Running CVIP-FEPC 761 12.3.2.2 Creating a New Project 761 12.3.2.3 Entering Classes in CVIP-FEPC 763 12.3.2.4 Adding Images and Associated Classes 763
xii Contents
12.3.2.5 Applying Feature Extraction and Pattern Classification.... 764 12.3.2.6 Running the Test 766 12.3.2.7 Result File 766
12.3.3 Application Development Example with Veterinary Thermographic Images 770 12.3.3.1 Introduction and Overview 770 12.3.3.2 Experiments 770 12.3.3.3 Results 775 12.3.3.4 Conclusion 775
12.4 Skin Lesion Classification Using Relative Color Features 775 12.4.1 Introduction and Project Overview 775 12.4.2 Materials and Methods 776
12.4.2.1 Image Database 776 12.4.2.2 Creation of Relative Color Images 776 12.4.2.3 Segmentation and Morphological Filtering 777 12.4.2.4 Feature Extraction 777 12.4.2.5 Lesion and Object Feature Spaces 779 12.4.2.6 Establishing Statistical Models 779
12.4.3 Experiments and Data Analysis 780 12.4.3.1 Lesion Feature Space 781 12.4.3.2 Object Feature Space 783
12.4.4 Conclusions 785 12.5 Automatic Segmentation of Blood Vessels in Retinal Images 786
12.5.1 Introduction and Overview 786 12.5.2 Materials and Methods 787 12.5.3 Results 792 12.5.4 Postprocessing with Hough Transform and Edge Linking 794 12.5.5 Conclusion 794
12.6 Classification of Land Types from Satellite Images Using Quadratic Discriminant Analysis and Multilayer Perceptrons 795 12.6.1 Introduction and Overview 795
12.6.2 Data Reduction and Feature Extraction 797 12.6.3 Object Classification 799 12.6.4 Results 800 12.6.5 Conclusion 801 12.6.6 Acknowledgments 803
12.7 Watershed-Based Approach to Skin Lesion Border Segmentation 803 12.7.1 Introduction 803 12.7.2 Materials and Methods 803 12.7.3 Experiments, Results, and Conclusions 809
12.8 Faint Line Defect Detection in Microdisplay (CCD) Elements 811 12.8.1 Introduction and Project Overview 811 12.8.2 Design Methodology 811 12.8.3 Line Detection Algorithm 812
12.8.3.1 Preprocessing 812 12.8.3.2 Edge Detection 814 12.8.3.3 Analysis of the Hough Space 816
12.8.4 Results and Discussion 819 12.8.5 Summary and Conclusion 820
Contents xiii
12.9 Melanoma and Seborrheic Keratosis Differentiation Using Texture Features 820 12.9.1 Introduction and Overview 820 12.9.2 Materials and Methods 821 12.9.3 Texture Analysis Experiments 823 12.9.4 Results and Discussion 830 12.9.5 Conclusion 830 12.9.6 Acknowledgments 831
12.10 Compression of Color Skin Tumor Images with Vector Quantization 831 12.10.1 Introduction and Project Overview 831 12.10.2 Materials and Methods 832
12.10.2.1 Compression Schemes 832 12.10.2.2 Subjective Evaluation of the Images 833