Optimisation in Signal and Image Processing

Patrick Siarry is a Professor of Automatics and Informatics at the University of Paris-Est Créteil, where he leads the Image and Signal Processing team in the Laboratoire Images, Signaux et Systèmes Intelligents - LiSSi.

• Introduction xiii Chapter 1. Modeling and Optimization in Image Analysis 1Jean Louchet1.1. Modeling at the source of image analysis and synthesis 11.2. From image synthesis to analysis 21.3. Scene geometric modeling and image synthesis 31.4. Direct model inversion and the Hough transform 41.5. Optimization and physical modeling 91.6. Conclusion 121.7. Acknowledgements 131.8. Bibliography 13Chapter 2. Artificial Evolution and the Parisian Approach. Applications in the Processing of Signals and Images 15Pierre Collet and Jean Louchet2.1. Introduction 152.2. The Parisian approach for evolutionary algorithms 152.3. Applying the Parisian approach to inverse IFS problems 172.4. Results obtained on the inverse problems of IFS 202.5. Conclusion on the usage of the Parisian approach for inverse IFS problems 222.6. Collective representation: the Parisian approach and the Fly algorithm 232.7. Conclusion 402.8. Acknowledgements 412.9.Bibliography 41Chapter 3. Wavelets and Fractals for Signal and Image Analysis 45Abdeldjalil Ouahabi and Djedjiga Ait Aouit3.1. Introduction 453.2. Some general points on fractals 463.3. Multifractal analysis of signals 543.4. Distribution of singularities based on wavelets 603.5. Experiments 703.6. Conclusion 763.7. Bibliography 76Chapter 4. Information Criteria: Examples of Applications in Signal and Image Processing 79Christian Oliver and Olivier Alata4.1. Introduction and context 794.2. Overview of the different criteria 814.3. The case of auto-regressive (AR) models 834.4. Applying the process to unsupervised clustering 954.5. Law approximation with the help of histograms 984.6. Other applications 1034.7. Conclusion 1064.8. Appendix 1064.9. Bibliography 107Chapter 5. Quadratic Programming and Machine Learning – Large Scale Problems and Sparsity 111Gaëlle Looslil, Stéphane Canu5.1. Introduction 1115.2. Learning processes and optimization 1125.3. From learning methods to quadratic programming 1175.4. Methods and resolution 1195.5. Experiments 1285.6. Conclusion 1325.7. Bibliography 133Chapter 6. Probabilistic Modeling of Policies and Application to Optimal Sensor Management 137Frédéric Dambreville, Francis Celeste and Cécile Simonin6.1. Continuum, a path toward oblivion 1376.2. The cross-entropy (CE) method 1386.3. Examples of implementation of CE for surveillance 1466.4. Example of implementation of CE for exploration 1536.5. Optimal control under partial observation 1586.6. Conclusion 1666.7. Bibliography 166Chapter 7. Optimizing Emissions for Tracking and Pursuit of Mobile Targets 169Jean-Pierre Le Cadre7.1. Introduction 1697.2. Elementary modeling of the problem (deterministic case) 1707.3. Application to the optimization of emissions (deterministic case) 1757.4. The case of a target with a Markov trajectory 1817.5. Conclusion 1897.6. Appendix: monotonous functional matrices 1897.7. Bibliography 192Chapter 8. Bayesian Inference and Markov Models 195Christophe Collet8.1. Introduction and application framework 1958.2. Detection, segmentation and classification 1968.3. General modeling 1998.4. Segmentation using the causal-in-scale Markov model 2018.5. Segmentation into three classes 2038.6. The classification of objects 2068.7. The classification of seabeds 2128.8. Conclusion and perspectives 2148.9. Bibliography 215Chapter 9. The Use of Hidden Markov Models for Image Recognition: Learning with Artificial Ants, Genetic Algorithms and Particle Swarm Optimization 219Sébastien Aupetit, Nicolas Monmarchè and Mohamed Slimane9.1. Introduction 2199.2. Hidden Markov models (HMMs) 2209.3. Using metaheuristics to learn HMMs 2239.4. Description, parameter setting and evaluation of the six approaches that are used to train HMMs 2269.5. Conclusion 2409.6. Bibliography 240Chapter 10. Biological Metaheuristics for Road Sign Detection 245Guillaume Dutilleux and Pierre Charbonnier10.1. Introduction 24510.2. Relationship to existing works 24610.3. Template and deformations 24810.4. Estimation problem 24810.5. Three biological metaheuristics 25210.6. Experimental results 25910.7. Conclusion 26510.8. Bibliography 266Chapter 11. Metaheuristics for Continuous Variables. The Registration of Retinal Angiogram Images 269Johann Drèo, Jean-Claude Nunes and Patrick Siarry11.1. Introduction 26911.2. Metaheuristics for difficult optimization problems 27011.3. Image registration of retinal angiograms 27511.4. Optimizing the image registration process 27911.5. Results 28811.6. Analysis of the results 29511.7. Conclusion 29611.8. Acknowledgements 29611.9. Bibliography 296Chapter 12. Joint Estimation of the Dynamics and Shape of Physiological Signals through Genetic Algorithms 301Amine Naït-Ali and Patrick Siarry12.1. Introduction 30112.2. Brainstem Auditory Evoked Potentials (BAEPs) 30212.3. Processing BAEPs 30312.4. Genetic algorithms 30512.5. BAEP dynamics 30712.6. The non-stationarity of the shape of the BAEPs 32412.7. Conclusion 32712.8. Bibliography 327Chapter 13. Using Interactive Evolutionary Algorithms to Help Fit Cochlear Implants 329Pierre Collet, Pierrick Legrand, Claire Bourgeois-République, Vincent Péan and Bruno Frachet13.1. Introduction 32913.2. Choosing an optimization algorithm 33313.3. Adapting an evolutionary algorithm to the interactive fitting of cochlear implants 33513.4. Evaluation 33813.5. Experiments 33913.6. Medical issues which were raised during the experiments 35013.7. Algorithmic conclusions for patient A 35213.8. Conclusion 35413.9. Bibliography 354List of Authors 357Index 359