●Preface
Acknowledgments
1 Information Theory Basics
1.1 Entropy
1.2 Relative Entropy and tual Information
1.3 Decomposition of tual Information
1.4 Inequalities
1.4.1 Jensen's Inequality
1.4.2 Log-sum Inequality
1.4.3 Jensen-Shannon Inequality
1.4.4 Data Processing Inequality
1.5 Entropy Rate
1.6 Entropy and Coding
1.7 Continuous Channel
1.8 Information Bottleneck Method
1.9 f-Divergences
1.10 Generalized Entropies
1.11 The Similarity Metric
2 Image Registration
2.1 The Registration Pipeline
2.1.1 Spatial Transform
2.1.2 Interpolation
2.1.3 Metric
2.1.4 Optimization
2.2 Similarity Metrics based on Shannon's Information Measures
2.2.1 Information Channel
2.2.2 Joint Entropy
2.2.3 tual Information
2.2.4 Normalized Measures
2.3 Probability Density Function Estimation
2.3.1 Histogram Estimation
2.3.2 Parzen Window Estimation
2.3.3 Entropic Spanning Graphs
2.4 High-dimensional Information Measures Including Spatial Information
2.5 Image Registration based on f-divergences
2.6 Similarity Measures based on Generalized Entropies
2.7 Measures based on The Similarity Metric
2.8 Image Fusion
2.8.1 Communication Channel
2.8.2 Specific Information
2.8.3 Fusion Criteria
2.8.4 Visualization
3 Image Segmentation
3.1 Maximum Entropy Thresholding
3.1.1 Entropy
3.1.2 Relative Entropy
3.2 Thresholding Considering Spatial Information
3.2.1 Grey-level Co-occurrence Matrix
3.2.2 Minimum Spatial Entropy Thresholding
3.2.3 Excess Entropy
3.3 Evolving Curves
3.4 Information Bottleneck Method for Image Segmentation
3.4.1 Split-and-Merge Algorithm
3.4.2 Histogram Clustering
3.4.3 Registration-based Segmentation
4 Video Key Frame Selection
4.1 Related Work and First IT-based Approaches
4.2 Key Frame Selection based on Jensen-Shannon Divergence and Jensen-Renyi Divergence
4.2.1 Jensen-Renyi Divergence
4.2.2 The Core Computational Mechanism
4.2.3 Locating Shots,Subshots,and Key Frames
4.3 Key Frame Selection Techniques Using Tsallis tual Information and Jensen-Tsallis Divergence for Shots with Hard Cuts
4.3.1 tual Information-based Similarity between Frames
4.3.2 Jensen-Tsallis-based Similarity between Frames
4.3.3 Keyframe Selection
4.4 Experimental Results
4.4.1 Results on JS and JR-based Methods
4.4.2 Results on JT and TMI Driven Techniques
4.5 Conclusion
5 Informational Aesthetics Measures
5.1 Introduction
5.2 Origins and Related Work
5.3 Global Aesthetic Measures
5.3.1 Shannon's Perspective
5.3.2 Kolmogorov's Perspective
5.3.3 Zurek's Perspective
5.4 Compositional Aesthetic Measures
5.4.1 Order as Self-Similarity
5.4.2 Interpreting Bense's Channel
5.5 Informational Analysis of Van Gogh's Periods
5.6 Towards Auvers Period: Evolution of van Gogh's Style
5.6.1 Randomness
5.6.2 Structural com plexity
5.6.3 Artistic Analysis
5.7 Color and Regional Information
A Digital-Image-Palette
Bibliography
Authors' Biographies

信息論方法廣泛應用於工程、物理、遺傳學、神經科學等科學領域，在圖像處理中也逐漸成為有用的工具。本書中介紹了信息論的基本概念，以及其如何在配準、分割、視頻處理和計算美學等圖像處理領域中應用。本書提出的一些方法，例如將互信息應用於圖像配準，是圖像處理的近期新技術。本書強調了信息論方法的共性方面，並以統一的方式介紹它們，以便向讀者表明信息論方法可以幫助解決圖像處理中的哪些問題，提供哪些特定工具，以及如何應用它們。本書可供圖像處理以及相關領域（例如計算機圖形和可視化）的學生和技術人員學習閱讀，IT領域學生和從業人員都會對了解這些應用感興趣。