Muutke küpsiste eelistusi

E-raamat: Evolutionary Computer Vision: The First Footprints

  • Formaat: PDF+DRM
  • Sari: Natural Computing Series
  • Ilmumisaeg: 28-Sep-2016
  • Kirjastus: Springer-Verlag Berlin and Heidelberg GmbH & Co. K
  • Keel: eng
  • ISBN-13: 9783662436936
Teised raamatud teemal:
  • Formaat - PDF+DRM
  • Hind: 55,56 €*
  • * hind on lõplik, st. muud allahindlused enam ei rakendu
  • Lisa ostukorvi
  • Lisa soovinimekirja
  • See e-raamat on mõeldud ainult isiklikuks kasutamiseks. E-raamatuid ei saa tagastada.
Evolutionary Computer Vision: The First FootprintsSuurem pilt
  • Formaat: PDF+DRM
  • Sari: Natural Computing Series
  • Ilmumisaeg: 28-Sep-2016
  • Kirjastus: Springer-Verlag Berlin and Heidelberg GmbH & Co. K
  • Keel: eng
  • ISBN-13: 9783662436936
Teised raamatud teemal:

DRM piirangud

  • Kopeerimine (copy/paste):

    ei ole lubatud

  • Printimine:

    ei ole lubatud

  • Kasutamine:

    Digitaalõiguste kaitse (DRM)
    Kirjastus on väljastanud selle e-raamatu krüpteeritud kujul, mis tähendab, et selle lugemiseks peate installeerima spetsiaalse tarkvara. Samuti peate looma endale  Adobe ID Rohkem infot siin. E-raamatut saab lugeda 1 kasutaja ning alla laadida kuni 6'de seadmesse (kõik autoriseeritud sama Adobe ID-ga).

    Vajalik tarkvara
    Mobiilsetes seadmetes (telefon või tahvelarvuti) lugemiseks peate installeerima selle tasuta rakenduse: PocketBook Reader (iOS / Android)

    PC või Mac seadmes lugemiseks peate installima Adobe Digital Editionsi (Seeon tasuta rakendus spetsiaalselt e-raamatute lugemiseks. Seda ei tohi segamini ajada Adober Reader'iga, mis tõenäoliselt on juba teie arvutisse installeeritud )

    Seda e-raamatut ei saa lugeda Amazon Kindle's. 

This book explains the theory and application of evolutionary computer vision, a new paradigm where challenging vision problems can be approached using the techniques of evolutionary computing. This methodology achieves excellent results for defining fitness functions and representations for problems by merging evolutionary computation with mathematical optimization to produce automatic creation of emerging visual behaviors.

In the first part of the book the author surveys the literature in concise form, defines the relevant terminology, and offers historical and philosophical motivations for the key research problems in the field. For researchers from the computer vision community, he offers a simple introduction to the evolutionary computing paradigm. The second part of the book focuses on implementing evolutionary algorithms that solve given problems using working programs in the major fields of low-, intermediate- and high-level computer vision.

This book will be ofvalue to researchers, engineers, and students in the fields of computer vision, evolutionary computing, robotics, biologically inspired mechatronics, electronics engineering, control, and artificial intelligence.

Arvustused

Olagues book aims to lay the foundations of this interdisciplinary research topic, concentrating particularly on goal-driven vision. This book is very pleasant to read, clear and well explained. Dr. Olague dots the text with original viewpoints, historical perspectives, and philosophical arguments. this excellent book is suitable for a large audience: beginners will find all the important features to start with, while specialists will find food for thought. (Evelyne Lutton, Genetic Programming and Evolvable Machines, 2017)

Muu info

"The book is written in a concise and complete manner ... . It will guide you through a new interdisciplinary field where the 3D modeling of computer vision is achieved using theoretical methods with elegant mathematics to applications with exciting intellectual results." (Long Quan, The Hong Kong University of Science and Technology) "This superb book, written by a leading researcher in evolutionary computing and computer vision, represents pioneering work where the principles of mathematical optimization are merged with the paradigm of artificial evolution in an original and productive way." (Marco Tomassini, Universite de Lausanne) "Excellent text provides a rigorous and scholarly account of evolutionary computing methods as a compelling way to approach computer vision ... . I expect this to be an essential source for students and researchers in evolutionary computing. For others, it may provide thought-provoking alternatives to mainstream computer vision techniques." (Steven M. LaValle, University of Illinois)
Part I Introduction
1 Introduction
3(8)
1.1 The Image Formation Process
3(2)
1.2 Motivation for Writing This Book
5(2)
1.3 Organization of the Book
7(4)
Part II Basics
2 Vision and Evolution: State of the Art
11(58)
2.1 What Is Vision?
11(11)
2.1.1 Brief History of Vision in Art
13(2)
2.1.2 Brief History of Geometry for Visual Representation
15(3)
2.1.3 Brief History of Photography and Photogrammetry
18(4)
2.2 Computer Vision
22(8)
2.2.1 Main Paradigms for Vision Understanding
22(2)
2.2.2 David Marr's Philosophy
24(2)
2.2.3 Goal-Driven Vision
26(4)
2.3 Evolution, Purpose and Teleology
30(9)
2.4 Evolutionary Computer Vision
39(2)
2.5 Computer Vision Applications
41(1)
2.6 Typical Hardware and Software Requirements
42(1)
2.7 Main Subjects Where EC Has Been Applied
43(5)
2.7.1 Early Visual Processing
44(1)
2.7.2 Intermediate Visual Processing
44(2)
2.7.3 High-Level Vision
46(1)
2.7.4 Others
47(1)
2.8 Conclusions on the State of the Art
48(21)
2.8.1 Research and Application Opportunities
48(1)
2.8.2 Where to Publish and Look for Information
48(5)
2.8.3 The Challenge of Being Human Competitive
53(2)
References
55(14)
3 Evolutionary Computing
69(74)
3.1 Introduction
69(4)
3.2 Evolutionary Computing as an Alternative to Optimization and Learning
73(11)
3.2.1 Mathematical Optimization
74(3)
3.2.2 Basic Terminology About Optimum
77(4)
3.2.3 Convex Optimization and Least Squares
81(3)
3.3 The Classical Evolutionary Algorithm
84(11)
3.3.1 Basic Components and Principles
85(2)
3.3.2 Evolutionary Computing as an Approach to Problem Solving
87(3)
3.3.3 Function as the Fundamental Concept in Artificial Evolution
90(2)
3.3.4 Problem Representation
92(3)
3.3.5 Fitness Function
95(1)
3.4 Genetic Algorithms
95(21)
3.4.1 Schema Theorem
102(3)
3.4.2 Between Chance and Determinism in Evolutionary Computing
105(2)
3.4.3 Genetic Operators and Evolutionary Algorithms
107(7)
3.4.4 Methods of Selection and Population Replacement
114(2)
3.5 Hierarchical Genetic Algorithms and Genetic Programming
116(3)
3.5.1 Basic GP Algorithm and Tree Representation
117(1)
3.5.2 GP as a Tool for Knowledge Discovery
117(1)
3.5.3 Variants of GP: Linear, Cartesian and Developmental Tree-Based GP
118(1)
3.6 Selected Topics
119(11)
3.6.1 Multiobjective Optimization
119(2)
3.6.2 Coevolution: Cooperative and Competitive
121(1)
3.6.3 Diversity and Species
122(1)
3.6.4 Differential Evolution
123(2)
3.6.5 Covariance Matrix Adaptation--ES
125(1)
3.6.6 Other Algorithms: PSO and ACO
126(1)
3.6.7 Artificial Life: Cellular Automata
127(2)
3.6.8 Evolutionary Robotics
129(1)
3.7 Conclusions
130(13)
References
133(10)
Part III Feature Location and Extraction
4 Accurate Modeling of Image Features Using Evolutionary Computing
143(50)
4.1 Introduction
143(2)
4.2 Modeling Corner Features
145(4)
4.2.1 Corner Morphology
146(1)
4.2.2 Corner Geometry
147(2)
4.2.3 Physical Properties of a Corner
149(1)
4.3 Corner Modeling
149(9)
4.3.1 Unit Step Edge Function Model
150(3)
4.3.2 L-corner Model
153(3)
4.3.3 Vertex Model
156(2)
4.4 Criteria for Accurate Corner Location
158(3)
4.5 Modeling Retro-reflective Targets
161(4)
4.5.1 Types of Retro-reflective Targets
161(1)
4.5.2 Overview of Previous Proposals
162(2)
4.5.3 Common Distortions in a Retro-reflective Target
164(1)
4.6 Analytical Model of a Retro-reflective Target
165(3)
4.7 Modeling of Data and Multidimensional Optimization
168(2)
4.7.1 Modeling Corners and Targets as an Optimization Problem
169(1)
4.8 Experimental Results
170(13)
4.9 Conclusions
183(10)
References
189(4)
5 Evolutionary Synthesis of Interest Point Detectors Through Genetic Programming
193(48)
5.1 Introduction
193(5)
5.2 Related Work
198(2)
5.3 Performance Criteria
200(5)
5.3.1 The Repeatability Rate
200(1)
5.3.2 Point Dispersion
201(1)
5.3.3 Information Content
202(1)
5.3.4 Holder Descriptor
203(2)
5.4 Evolving Interest Point Operators with Genetic Programming
205(2)
5.5 Design of Interest Operators Using a Single Objective Function
207(3)
5.5.1 Search Space
207(2)
5.5.2 Evaluation Function
209(1)
5.6 Design of Interest Operators with a Multiobjective Approach
210(3)
5.6.1 Improved Strength Pareto Evolutionary Algorithm
211(1)
5.6.2 Objective Functions and Search Space
211(2)
5.7 Experimental Results
213(21)
5.7.1 Single Objective Approach
213(3)
5.7.2 Multiobjective Approach
216(2)
5.7.3 Stability vs. Point Dispersion
218(6)
5.7.4 Stability vs. Information Content
224(3)
5.7.5 Point Dispersion vs. Information Content
227(3)
5.7.6 Stability, Point Dispersion, and Information Content
230(1)
5.7.7 Computational Cost
230(4)
5.8 Summary and Conclusions
234(7)
References
235(6)
Part IV 3D Computer Vision
6 The Honeybee Search Algorithm
241(32)
6.1 Introduction
241(3)
6.2 Parisian Evolution: Cooperative Coevolution Through Honeybee Search
244(1)
6.3 Problem Statement
245(3)
6.4 The Honeybee Dance Language
248(3)
6.5 The Honeybee Search Algorithm
251(6)
6.5.1 Fitness Function Evaluation
254(1)
6.5.2 Evolutionary Search Operators: Crossover, Mutation, and Sharing
255(2)
6.6 Experimental Results: Tuning the Algorithm
257(7)
6.6.1 Tests with the Mutation Operator
257(4)
6.6.2 Tests with the Crossover Operator
261(1)
6.6.3 Tests with the Sharing Operator
262(2)
6.7 Experimental Results: Testing with Standard Images
264(5)
6.8 Conclusions
269(4)
References
271(2)
7 Multiobjective Sensor Planning for Accurate Reconstruction
273(56)
7.1 Introduction
273(2)
7.2 Related Work
275(5)
7.2.1 Main Research Areas for Sensor Planning
276(2)
7.2.2 Photogrammetric Network Design
278(2)
7.3 Multiobjective Problem Design
280(4)
7.3.1 Multicriteria Optimization
282(1)
7.3.2 Multicriteria Decision Analysis
283(1)
7.4 Multiobjective Sensor Planning
284(12)
7.4.1 Problem Statement
284(1)
7.4.2 Accurate 3D Reconstruction
285(2)
7.4.3 Limited Error Propagation
287(4)
7.4.4 Criterion for Optimal Uncertainty
291(1)
7.4.5 Visualization of Uncertainty
292(1)
7.4.6 Efficient Robot Motion
293(1)
7.4.7 Computational Cost
294(1)
7.4.8 Pareto Optimal Sensing Strategies
295(1)
7.5 Evolutionary-Based Optimization
296(9)
7.5.1 Problem Variables and Representation
296(1)
7.5.2 Genotype-to-Phenotype Transformation
297(1)
7.5.3 Constraint Handling
298(1)
7.5.4 Recombination and Mutation
299(2)
7.5.5 The Evolutionary Engine
301(4)
7.6 Experimental Results
305(7)
7.6.1 Fixed Size Camera Networks
305(2)
7.6.2 Varying the Size of Camera Networks
307(1)
7.6.3 Visualization of the Complete Landscape
308(2)
7.6.4 Measuring a Plane
310(1)
7.6.5 Measuring a Two-plane Object
311(1)
7.6.6 Measuring a Complex Object
311(1)
7.6.7 Measuring a Real-world Object
312(1)
7.7 Conclusions
312(17)
References
323(6)
Part V Learning and Recognition
8 Evolutionary Visual Learning with Linear Genetic Programming
329(20)
8.1 Introduction
329(3)
8.1.1 Genetic Programming for Object Recognition
330(1)
8.1.2 Genetic Programming for Multiclass Object Recognition
331(1)
8.2 Second-Order Statistical Methods for Texture Analysis
332(1)
8.3 Location and Synthesis of Texture Features
333(4)
8.3.1 Linear Genetic Programming for Visual Learning
333(4)
8.4 Experiments with the Improved CALTECH Image Database
337(7)
8.4.1 Multiclass Object Recognition Database
337(1)
8.4.2 Experiment Using Three Classes
337(3)
8.4.3 Experiment Using Five Classes
340(4)
8.5 Conclusions
344(5)
References
347(2)
9 Evolutionary Synthesis of Feature Descriptor Operators with Genetic Programming
349(36)
9.1 Introduction
349(7)
9.1.1 Motivation and Problem Statement
351(1)
9.1.2 Research Contributions
352(1)
9.1.3 Related Work on Local Descriptors
353(3)
9.2 Evolving SIFT Operators with Genetic Programming
356(6)
9.2.1 Structure Representation, Search Space and Genetic Operations
358(2)
9.2.2 Fitness Function and the F-Measure
360(1)
9.2.3 Initialization, GP Parameters and Solution Designation
361(1)
9.3 Performance Evaluation
362(3)
9.4 Results
365(10)
9.4.1 Learning and Testing SIFT-RDGP Operators Through GP
366(2)
9.4.2 Experimental Evaluation of Local Descriptors
368(4)
9.4.3 Object Recognition Application
372(3)
9.5 Conclusions
375(10)
References
379(6)
Part VI Finale
10 Summary and Conclusions
385(6)
10.1 Summary
385(3)
10.1.1 Contributions and Feasibility of ECV
386(1)
10.1.2 A Human-Competitive Perspective
387(1)
10.1.3 Possible Extensions and Future Research
387(1)
10.2 Conclusions
388(3)
A Camera Calibration and Stereoscopic Vision
391(10)
A.1 The Projection Matrix
391(2)
A.2 System of Equations
393(1)
A.3 Camera Calibration
394(2)
A.4 The Epipolar Relationship Right-Left
396(5)
A.4.1 The Epipolar Constraint
397(1)
A.4.2 The Essential Matrix
398(1)
A.4.3 The Fundamental Matrix
399(2)
References 401(2)
Glossary 403(4)
Index 407
Gustavo Olague received his Ph.D. in Computer Vision, Graphics and Robotics from INPG (Institut Polytechnique de Grenoble) and INRIA (Institut National de Recherche en Informatique et Automatique) in France. He is a Professor in the Dept. of Computer Science at CICESE (Centro de Investigación Científica y de Educación Superior de Ensenada) in Mexico, and the Director of its EvoVisión Research Team. He is also an Adjoint Professor of Engineering at UACH (Universidad Autonóma de Chihuahua). He has authored over 100 conference proceedings papers and journal articles, he coedited two special issues in Pattern Recognition Letters and Evolutionary Computation, and he served as cochair of the Real-World Applications track at the main international evolutionary computing conference, GECCO (ACM SIGEVO Genetic and Evolutionary Computation Conference). Prof. Olague has received numerous distinctions, among them the Talbert Abrams Award presented by the American Society for Photogrammetry and Remote Sensing (ASPRS) for authorship and recording of current and historical engineering and scientific developments in photogrammetry; Best Paper awards at major conferences such as GECCO, EvoIASP (European Workshop on Evolutionary Computation in Image Analysis, Signal Processing and Pattern Recognition) and EvoHOT (European Workshop on Evolutionary Hardware Optimization); and twice the Bronze Medal at the Humies (GECCO award for Human-Competitive results produced by genetic and evolutionary computation). His main research interests are evolutionary computing and computer vision.
Lisainfo e-raamatute kohta Lisainfo e-raamatute kohta
Püsilink: https://www.kriso.ee/db/97836624369362e.html
Märksõnad: