Muutke küpsiste eelistusi

Stochastic Geometry for Image Analysis [Kõva köide]

4.00/5 (1 hinnangut Goodreads-ist)
Edited by
  • Formaat: Hardback, 384 pages, kõrgus x laius x paksus: 240x163x26 mm, kaal: 671 g
  • Ilmumisaeg: 25-Nov-2011
  • Kirjastus: ISTE Ltd and John Wiley & Sons Inc
  • ISBN-10: 1848212402
  • ISBN-13: 9781848212404
Teised raamatud teemal:
Stochastic Geometry for Image AnalysisSuurem pilt
  • Formaat: Hardback, 384 pages, kõrgus x laius x paksus: 240x163x26 mm, kaal: 671 g
  • Ilmumisaeg: 25-Nov-2011
  • Kirjastus: ISTE Ltd and John Wiley & Sons Inc
  • ISBN-10: 1848212402
  • ISBN-13: 9781848212404
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9781848212404.html
Märksõnad:
This book develops the stochastic geometry framework for image analysis purpose. Two main frameworks are  described: marked point process and random closed sets models. We derive the main issues for defining an appropriate model. The algorithms for sampling and optimizing the models as well as for estimating parameters are reviewed.  Numerous applications, covering remote sensing images, biological and medical imaging, are detailed.  This book provides all the necessary tools for developing an image analysis application based on modern stochastic modeling.
Chapter 1 Introduction
1(10)
X. Descombes
Chapter 2 Marked Point Processes for Object Detection
11(18)
X. Descombes
2.1 Principal definitions
11(4)
2.2 Density of a point process
15(6)
2.3 Marked point processes
21(1)
2.4 Point processes and image analysis
22(7)
2.4.1 Bayesian versus non-Bayesian
22(4)
2.4.2 A priori versus reference measure
26(3)
Chapter 3 Random Sets for Texture Analysis
29(36)
C. Lantuejoul
M. Schmitt
3.1 Introduction
29(4)
3.2 Random sets
33(9)
3.2.1 Insufficiency of the spatial law
33(1)
3.2.2 Introduction of a topological context
34(2)
3.2.3 The theory of random closed sets (RACS)
36(2)
3.2.4 Some examples
38(3)
3.2.5 Stationarity and isotropy
41(1)
3.3 Some geostatistical aspects
42(9)
3.3.1 The ergodicity assumption
42(1)
3.3.2 Inference of the DF of a stationary ergodic RACS
42(1)
3.3.2.1 Construction of the estimator
43(1)
3.3.2.2 On sampling
44(3)
3.3.3 Individual analysis of objects
47(4)
3.4 Some morphological aspects
51(10)
3.4.1 Geometric interpretation
52(1)
3.4.1.1 Point
52(1)
3.4.1.2 Pair of points
53(1)
3.4.1.3 Segment
54(1)
3.4.1.4 Ball
55(2)
3.4.2 Filtering
57(1)
3.4.2.1 Opening and closing
57(3)
3.4.2.2 Sequential alternate filtering
60(1)
3.5 Appendix: demonstration of Miles' formulae for the Boolean model
61(4)
Chapter 4 Simulation and Optimization
65(48)
F. Lafarge
X. Descombes
E. Zhizhina
R. Minlos
4.1 Discrete simulations: Markov chain Monte Carlo algorithms
66(25)
4.1.1 Irreducibility, recurrence, and ergodicity
67(1)
4.1.1.1 Definitions
67(1)
4.1.1.2 Stationarity
68(1)
4.1.1.3 Convergence
69(1)
4.1.1.4 Irreducibility
69(1)
4.1.1.5 Aperiodicity
70(1)
4.1.1.6 Harris recurrence
70(1)
4.1.1.7 Ergodicity
71(1)
4.1.1.8 Geometric ergodicity
72(1)
4.1.1.9 Central limit theorem
72(1)
4.1.2 Metropolis-Hastings algorithm
73(3)
4.1.3 Dimensional jumps
76(1)
4.1.3.1 Mixture of kernels
77(2)
4.1.3.2 π-reversibility
79(2)
4.1.4 Standard proposition kernels
81(1)
4.1.4.1 Simple perturbations
81(1)
4.1.4.2 Model switch
81(3)
4.1.4.3 Birth and death
84(3)
4.1.5 Specific proposition kernels
87(1)
4.1.5.1 Creating complex transitions from standard transitions
88(1)
4.1.5.2 Data-driven perturbations
89(1)
4.1.5.3 Perturbations directed by the current state
90(1)
4.1.5.4 Composition of kernels
90(1)
4.2 Continuous simulations
91(14)
4.2.1 Diffusion algorithm
91(4)
4.2.2 Birth and death algorithm
95(2)
4.2.3 Muliple births and deaths algorithm
97(1)
4.2.3.1 Convergence of the distributions
98(2)
4.2.3.2 Birth and death process
100(1)
4.2.4 Discrete approximation
100(2)
4.2.4.1 Acceleration of the multiple births and deaths algorithm
102(3)
4.3 Mixed simulations
105(1)
4.3.1 Jump process
105(1)
4.3.2 Diffusion process
105(1)
4.3.3 Coordination of jumps and diffusions
106(1)
4.4 Simulated annealing
106(7)
4.4.1 Cooling schedule
107(1)
4.4.2 Initial temperature T0
108(1)
4.4.3 Logarithmic decrease
109(1)
4.4.4 Geometric decrease
109(1)
4.4.5 Adaptive reduction
110(2)
4.4.6 Stopping criterion/final temperature
112(1)
Chapter 5 Parametric Inference for Marked Point Processes in Image Analysis
113(48)
R. Stoica
F. Chatelain
M. Sigelle
5.1 Introduction
113(4)
5.2 First question: what and where are the objects in the image?
117(12)
5.3 Second question: what are the parameters of the point process that models the objects observed in the image?
129(29)
5.3.1 Complete data
130(1)
5.3.1.1 Maximum likelihood
130(11)
5.3.1.2 Maximum pseudolikelihood
141(10)
5.3.2 Incomplete data: EM algorithm
151(7)
5.4 Conclusion and perspectives
158(1)
5.5 Acknowledgments
159(2)
Chapter 6 How to Set Up a Point Process?
161(18)
X. Descombes
6.1 From disks to polygons, via a discussion of segments
162(5)
6.2 From no overlap to alignment
167(5)
6.3 From the likelihood to a hypothesis test
172(4)
6.4 From Metropolis-Hastings to multiple births and deaths
176(3)
Chapter 7 Population Counting
179(70)
X. Descombes
7.1 Detection of Virchow-Robin spaces
180(12)
7.1.1 Data modeling
181(3)
7.1.2 Marked point process
184(3)
7.1.3 Reversible jump MCMC algorithm
187(3)
7.1.4 Results
190(2)
7.2 Evaluation of forestry resources
192(15)
7.2.1 2D model
193(1)
7.2.1.1 Prior
193(4)
7.2.1.2 Data term
197(2)
7.2.1.3 Optimization
199(2)
7.2.1.4 Results
201(4)
7.2.2 3D model
205(2)
7.2.2.1 Results
207(1)
7.3 Counting a population of flamingos
207(22)
7.3.1 Estimation of the flamingo color
213(4)
7.3.2 Simulation and optimization by multiple births and deaths
217(1)
7.3.3 Results
218(11)
7.4 Counting the boats at a port
229(20)
7.4.1 Initialization of the optimization algorithm
234(1)
7.4.1.1 Parameter γd
234(2)
7.4.1.2 Calibration of the d0 parameter
236(1)
7.4.2 Initial results
237(2)
7.4.3 Modification of the data energy
239(2)
7.4.3.1 First modification of the prior energy
241(4)
7.4.3.2 Second modification of the prior energy
245(4)
Chapter 8 Structure Extraction
249(38)
F. Lafarge
X. Descombes
8.1 Detection of the road network
250(12)
8.2 Extraction of building footprints
262(7)
8.3 Representation of natural textures
269(18)
8.3.1 Simple model
274(1)
8.3.1.1 Data term
275(3)
8.3.1.2 Sampling by jump diffusion
278(1)
8.3.1.3 Results
279(4)
8.3.2 Models with complex interactions
283(4)
Chapter 9 Shape Recognition
287(38)
F. Lafarge
C. Mallet
9.1 Modeling of a LIDAR signal
287(21)
9.1.1 Motivation
290(1)
9.1.2 Model library
291(2)
9.1.2.1 Energy formulation
293(4)
9.1.3 Sampling
297(1)
9.1.4 Results
298(1)
9.1.4.1 Simulated data
298(2)
9.1.4.2 Satellite data: large footprint waveforms
300(2)
9.1.4.3 Airborne data: small footprint waveforms
302(4)
9.1.4.4 Application to the classification of 3D point clouds
306(2)
9.2 3D reconstruction of buildings
308(17)
9.2.1 Library of 3D models
308(3)
9.2.2 Bayesian formulation
311(2)
9.2.2.1 Likelihood
313(1)
9.2.2.2 A priori
314(3)
9.2.3 Optimization
317(1)
9.2.4 Results and discussion
318(7)
Bibliography 325(16)
List of Authors 341(2)
Index 343
Xavier Descombes, Director of Research INRIA, EPI Ariana