Digital Image Analysis - Selected Techniques and Applications

Preface

About This Book

The Compact Disc

Acknowledgments

Contributors

Contents

List of Figures

List of Tables

Part I Mathematical Methods for Image Analysis

Introduction to Part I

1 Numerical Harmonic Analysis and Image Processing

1.1 Gabor Analysis and Digital Signal Processing

1.1.1 From Fourier to Gabor Expansions

1.1.2 Local Time-Frequency Analysis and Short-Time Fourier Transform

1.1.3 Fundamental Properties of Gabor Frames

1.1.4 Commutation Relations of the Gabor Frame Operator

1.1.5 Critical Sampling, Oversampling, and the Balian-Low Theorem

1.1.6 Wexler-Raz Duality Condition

1.1.7 Gabor Analysis on LCA Groups

1.1.8 Numerical Gabor Analysis

1.1.9 Image Representation and Gabor Analysis

1.2 Signal and Image Reconstruction

1.2.1 Notation

1.2.2 Signal Reconstruction and Frames

1.2.3 Numerical Methods for Signal Reconstruction

1.3 Examples and Applications

1.3.1 Object Boundary Recovery in Echocardiography

1.3.2 Image Reconstruction in Exploration Geophysics

1.3.3 Reconstruction of Missing Pixels in Images

2 Stochastic Shape Theory

2.1 Shape Analysis

2.2 Contour Line Parameterization

2.3 Deformable Templates

2.3.1 Stochastic Planar Deformation Processes

2.3.2 Gaussian Isotropic Random Planar Deformations

2.3.3 The Deformable Templates Model

2.3.4 Maximum Likelihood Classi.cation

2.4 The Wavelet Transform

2.4.1 Atomic Decompositions and Group Theory

2.4.2 Discrete Wavelets and Multiscale Analysis

2.4.3 Wavelet Packets

2.5 Wavelet Packet Descriptors

2.6 Global Nonlinear Optimization

2.6.1 Multilevel Single-Linkage Global Optimization

2.6.2 Implementation

3 Image Compression and Coding

3.1 Image Compression

3.1.1 Lossy Compression and Machine Vision

3.1.2 Multilevel Polynomial Interpolation

3.1.3 Enhancing the FBI Fingerprint Compression Standard

3.2 Multimedia Data Encryption

3.2.1 Symmetric Product Ciphers

3.2.2 Permutation by Chaotic Kolmogorov Flows

3.2.3 Substitution by AWC or SWB Generators

3.2.4 Security Considerations

3.2.5 Encryption Experiments

3.2.6 Encryption Summary

References

Part II Data Handling

Introduction to Part II

4 Parallel and Distributed Processing

4.1 Dealing with Large Remote Sensing Image Data Sets

4.1.1 Demands of Earth Observation

4.1.2 Processing Radar-Data of the Magellan Venus Probe

4.2 Parallel Radar Signal Processing

4.2.1 Parallelization Strategy

4.2.2 Evaluation of Parallelization Tools

4.2.3 Program Analysis and Parallelization

4.3 Parallel Radar Image Processing

4.3.1 Data Decomposition and Halo Handling

4.3.2 Dynamic Load Balancing and Communication Overloading

4.3.3 Performance Assessment

4.4 Distributed Processing

4.4.1 Front End

4.4.2 Back End

4.4.3 Broker

4.4.4 Experiences

5 Image Data Catalogs

5.1 Online Access to Remote Sensing Imagery

5.1.1 Remote Sensing Data Management

5.1.2 Image Data Information and Request System

5.1.3 Online Product Generation and Delivery

5.2 Content-Based Image Database Indexing and Retrieval

5.2.1 The Miniature Portrait Database

5.2.2 The Eigen Approach

5.2.3 Experiments

References

Part II Robust and Adaptive Image Understanding

Introduction to Part III

6 Graphs in Image Analysis

6.1 From Pixels to Graphs

6.1.1 Graphs in the Square Grid

6.1.2 Run Graphs

6.1.3 Area Voronoi Diagram

6.2 Graph Transformations in Image Analysis

6.2.1 Arrangements of Image Elements

6.2.2 Dual Graph Contraction

7 Hierarchies

7.1 Regular Image Pyramids

7.1.1 Structure

7.1.2 Contents

7.1.3 Processing

7.1.4 Fuzzy Curve Pyramid

7.2 Irregular Graph Pyramids

7.2.1 Computational Complexity

7.2.2 Irregular Pyramids by Hop.eld Networks

7.2.3 Equivalent Contraction Kernels

7.2.4 Extensions to 3D

8 Robust Methods

8.1 The Role of Robustness in Computer Vision

8.2 Parametric Models

8.2.1 Robust Estimation Methods

8.3 Robust Methods in Vision

8.3.1 Recover-and-Select Paradigm

8.3.2 Recover-and-Select applied to

9 Structural Object Recognition

9.1 2-D and 3-D Structural Features

9.2 Feature Selection

9.3 Matching Structural Descriptions

9.4 Reducing Search Complexity

9.5 Grouping and Indexing

9.5.1 Early Search Termination

9.6 Detection of Polymorphic Features

9.7 Polymorphic Grouping

9.8 Indexing and Matching

9.9 Polymorphic Features

9.10 3-D Object Recognition Example

9.10.1 The IDEAL System

9.10.2 Initial Structural Part Decomposition

9.10.3 Part Adjacency and Compatibility Graphs

9.10.4 Automatic Model Acquisition

9.10.5 Object Recognition from Appearances

9.10.6 Experiments

10 Machine Learning

10.1 What Is Machine Learning?

10.1.1 What Do Machine Learning Algorithms Need?

10.1.2 One Method Solves All? Use of Multistrategy

10.2 Methods

10.3 Operational

10.3.1 Discrimination and Classi.cation

10.3.2 Optimization and Search

10.3.3 Functional Relationship

10.3.4 Logical Operations

10.4 Object-Oriented Generalization

10.5 Generalized Logical Structures

10.5.1 Reformulation

10.5.2 Object-Oriented Implementation

10.6 Generalized Clustering Algorithms

10.6.1 Function Overloading

Conclusion

References

Part IV Information Fusion and Radiometric Models for Image Understanding

Introduction to Part IV

11 Information Fusion in Image Understanding

11.1 Active Fusion

11.2 Active Object Recognition

11.2.1 Related Research

11.3 Feature Space Active Recognition

11.3.1 Object Recognition in Parametric Eigenspace

11.3.2 Probability Distributions in Eigenspace

11.3.3 View Classi.cation and Pose Estimation

11.3.4 Information Integration

11.3.5 View Planning

11.3.6 The Complexity of the Algorithm

11.3.7 Experiments

11.3.8 A Counterexample for Conditional Independence in Equation ( 11.5)

11.3.9 Conclusion

11.4 Reinforcement Learning for Active Object Recognition

11.4.1 Adaptive Generation of Object Hypotheses

11.4.2 Learning Recognition Control

11.4.3 Experiments

11.4.4 Discussion and Outlook

11.5 Generic Active Object Recognition

11.5.1 Object Models

11.5.2 Recognition System

11.5.3 Hypothesis Generation

11.5.4 Visibility Space

11.5.5 Viewpoint Estimation

11.5.6 Viewpoints and Actions

11.5.7 Motion Planning

11.5.8 Object Hypotheses Fusion

11.5.9 Conclusion

12 Image Understanding Methods for Remote Sensing

12.1 Radiometric Models

12.2 Subpixel Analysis of Remotely Sensed Images

12.3 Segmentation of Remotely Sensed Images

12.4 Land-Cover Classi.cation

12.5 Information Fusion for Remote Sensing

References

Part V 3D Reconstruction

Introduction to Part V

13 Fundamentals

13.1 Image Acquisition Aspects

13.1.1 Video Cameras

13.1.2 Amateur Cameras with CCD Sensors

13.1.3 Analog Metric Cameras

13.1.4 Remote Sensing Scanners

13.1.5 Other Visual Sensor Systems

13.2 Perspective Transformation

13.3 Stereo Reconstruction

13.4 Bundle Block Con.gurations

13.5 From Points and Lines to Surfaces

13.5.1 Representation of Irregular Object Surfaces

13.5.2 Representation of Man-Made Objects

13.5.3 Hybrid Representation of Object Surfaces

14 Image Matching Strategies

14.1 Raster-Based Matching Techniques

14.1.1 Cross Correlation

14.1.2 Least Squares Matching

14.2 Feature-Based Matching Techniques

14.2.1 Feature Extraction

14.2.2 Matching Homologous Image Features

14.3 Hierarchical Feature Vector Matching (HFVM)

14.3.1 Feature Vector Matching (FVM)

14.3.2 Subpixel Matching

14.3.3 Consistency Check

14.3.4 Hierarchical Feature Vector Matching

15 Precise Photogrammetric Measurement: Location of Targets and Reconstruction of Object Surfaces

15.1 Automation in Photogrammetric Plotting

15.1.1 Automation of Inner Orientation

15.1.2 Automation of Outer Orientation

15.2 Location of Targets

15.2.1 Location of Circular Symmetric Targets by Intersection of Gradient Vectors

15.2.2 Location of Arbitrarily Shaped Targets

15.2.3 The OEEPE Test on Digital Aerial Triangulation

15.2.4 Deformation Analysis of Wooden Doors

15.3 A General Framework for Object Reconstruction

15.3.1 Hierarchical Object Reconstruction

15.3.2 Mathematical Formulation of the Object Models

15.3.3 Robust Hybrid Adjustment

15.3.4 DEM Generation for Topographic Mapping

15.4 Semiautomatic Building Extraction

15.4.1 Building Models

15.4.2 Interactive Determination of Approximations

15.4.3 Automatic Fine Reconstruction

15.5 State of Work

16 3D Navigation and Reconstruction

16.1 High Accurate Stereo Reconstruction of Naturally Textured Surfaces for Navigation and 3D- Modeling

16.1.1 Reconstruction of Arbitrary Shapes Using the Locus Method

16.1.2 Using the locus Method for Cavity Inspection

16.1.3 Stereo Reconstruction Using Remote Sensing Images

16.1.4 Stereo Reconstruction for Space Research

16.1.5 Operational Industrial Stereo Vision Systems

16.2 A Framework for Vision-Based Navigation

16.2.1 Vision Sensor Systems

16.2.2 Closed-Loop Solution for Autonomous Navigation

16.2.3 Risk Map Generation

16.2.4 Local Path Planning

16.2.5 Path Execution and Navigation on the DEM

16.2.6 Prototype Software for Closed-Loop Vehicle Navigation

16.2.7 Simulation Results

17 3D Object Sensing Using Rotating CCD Cameras

17.1 Concept of Image-Based Theodolite Measurement Systems

17.2 The Videometric Imaging System

17.2.1 The Purpose of the Videometric Imaging System

17.2.2 An Interactive Measurement System–A First Step

17.2.3 An Automatic System–A Second Step

17.3 Conversion of the Measurement System into a Robot System

17.4 Decision Making

17.5 Outlook

References

Index