Contents

Part I

Maïtine Bergounioux

Second-order decomposition model for image processing: numerical experimentation

1.1Introduction

1.2Presentation of the model

1.3Numerical aspects

1.3.1Discretized problem and algorithm

1.3.2Examples

1.3.3Initialization process

1.3.4Convergence

1.3.5Sensitivity with respect to sampling and quantification

1.3.6Sensitivity with respect to parameters

1.4Conclusion

Laurent Hoeltgen, Markus Mainberger, Sebastian Hoffmann, Joachim Weickert, Ching Hoo Tang, Simon Setzer, Daniel Johannsen, Frank Neumann, and Benjamin Doerr

Optimizing spatial and tonal data for PDE-based inpainting

2.1Introduction

2.2A review of PDE-based image compression

2.2.1Data optimization

2.2.2Finding good inpainting operators

2.2.3Storing the data

2.2.4Feature-based methods

2.2.5Fast algorithms and real-time aspects

2.2.6Hybrid image compression methods

2.2.7Modifications, extensions and applications

2.2.8Relations to other methods

2.3Inpainting with homogeneous diffusion

2.4Optimization strategies in 1D

2.4.1Optimal knots for interpolating convex functions

2.4.2Optimal knots for approximating convex functions

2.5Optimization strategies in 2D

2.5.1Optimizing spatial data

2.5.2Optimizing tonal data

2.6Extensions to other inpainting operators

2.6.1Optimizing spatial data

2.6.2Optimizing tonal data

2.7Summary and conclusions

Qian Xie and Anuj Srivastava

Image registration using phaseamplitude separation

3.1Introduction

3.1.1Current literature

3.1.2Our approach

3.2Definition of phaseamplitude components

3.2.1q-Map and amplitude distance

3.2.2Relative phase and image registration

3.3Properties of registration framework

3.4Gradient method for optimization over Γ

3.4.1Basis on

3.4.2Mean image and group-wise registration

3.5Experiments

3.5.1Pairwise image registration

3.5.2Registering multiple images

3.5.3Image classification

3.6Conclusion

Martin Eller and Massimo Fornasier

Rotation invariance in exemplar-based image inpainting

4.1Introduction to inpainting

4.1.1The inpainting problem

4.1.2Aims of this work

4.1.3Notation

4.2Rotation invariant image pattern recognition

4.2.1Patch error functions

4.2.2Circular harmonics basis

4.2.3Mutual angle detection algorithms

4.2.4Rotation invariant L2-error using the circular harmonics basis

4.2.5Rotation invariant gradient-based L2-errors and the CH-basis

4.3Rotation invariant exemplar-based inpainting

4.3.1Patch non-local means

4.3.2Patch non-local Poisson

4.3.3Numerical experiments

4.4Discussion and analysis

4.4.1Proof of convergence

4.4.2Analysis of E,T

4.4.3Conclusion and future perspectives

José A. Iglesias and Clemens Kirisits

Convective regularization for optical flow

5.1Introduction

5.2Model

5.2.1Convective acceleration

5.2.2Convective regularization

5.2.3Data term and contrast invariance

5.3Numerical solution

5.4Experiments

5.5Conclusion

Elena Beretta, Monika Muszkieta, Wolf Naetar, and Otmar Scherzer

A variational method for quantitative photoacoustic tomography with piecewise constant coefficients

6.1Quantitative photoacoustic tomography

6.1.1Introduction

6.1.2Contributions of this article

6.2Recovery of piecewise constant coefficients

6.3A MumfordShah-like functional for qPAT

6.3.1Existence of minimizers

6.3.2Approximation

6.3.3Minimization

6.4Implementation and numerical results

ASpecial functions of bounded variation and the SBV-compactness theorem

Martin Burger, Hendrik Dirks, and Lena Frerking

On optical flow models for variational motion estimation

7.1Introduction

7.2Models

7.2.1Variational models with gradient regularization

7.2.2Extension of the regularizer

7.2.3Bregman iterations

7.3Analysis

7.3.1Existence of minimizers

7.3.2Quantitative estimates

7.4Numerical solution

7.4.1Primaldual algorithm

7.4.2Discretization and parameters

7.5Results

7.5.1Error measures for velocity fields

7.5.2Evaluation

7.6Conclusion and outlook

7.6.1Mass preservation

7.6.2Higher dimensions

7.6.3Joint models

7.6.4Large displacements

Luca Calatroni, Chung Cao, Juan Carlos De los Reyes, Carola-Bibiane Schönlieb, and Tuomo Valkonen

Bilevel approaches for learning of variational imaging models

8.1Overview of learning in variational imaging

8.2The learning model and its analysis in function space

8.2.1The abstract model

8.2.2Existence and structure: L2-squared cost and fidelity

8.2.3Optimality conditions

8.3Numerical optimization of the learning problem

8.3.1Adjoint-based methods

8.3.2Dynamic sampling

8.4Learning the image model

8.4.1Total variation-type regularization

8.4.2Optimal parameter choice for TV-type regularization

8.5Learning the data model

8.5.1Variational noise models

8.5.2Single noise estimation

8.5.3Multiple noise estimation

8.6Conclusion and outlook

Part II

Piernicola Bettiol and Nathalie Khalil

Non-degenerate forms of the generalized EulerLagrange condition for state-constrained optimal control problems

9.1Introduction

9.2Main result

9.3Proof of Theorem 2.1

9.4Proof of Lemma 3.2

9.5Example

Piernicola Bettiol, Bernard Bonnard, Laetitia Giraldi, Pierre Martinon, and Jérémy Rouot

The Purcell three-link swimmer: some geometric and numerical aspects related to periodic optimal controls

10.1Introduction

10.2First- and second-order optimality conditions

10.3The Purcell three-link swimmer

10.3.1Mathematical model

10.4Local analysis for the three-link Purcell swimmer

10.4.1Computations of the nilpotent approximation

10.4.2Integration of extremal trajectories

10.5Numerical results

10.5.1Nilpotent approximation

10.5.2True mechanical system

10.5.3The Purcell swimmer in a round swimming pool

10.6Conclusions and future work

Zheng Chen and Yacine Chitour

Controllability of Keplerian motion with low-thrust control systems

11.1Introduction

11.2Notations and definitions

11.2.1Dynamics

11.2.2Study of the drift vector field in ?

11.2.3Admissible controlled trajectory of Σsat

11.2.4Controlled problems in ?

11.3Controllability

11.3.1Controllability for OTP

11.3.2Controllability for OIP

11.3.3Controllability for DOP

11.4Numerical examples

11.4.1A numerical example for OIP

11.4.2A numerical example for DOP

11.5Conclusion

11.6Appendix

Thierry Combot

Higher variational equation techniques for the integrability of homogeneous potentials

12.1Introduction: integrable systems

12.2An algebraic point of view

12.2.1Algebraic presentation of a Hamiltonian system

12.2.2First-order variational equations

12.2.3Differential Galois theory

12.3Introduction to MoralesRamis theorem

12.3.1The MoralesRamis theorem

12.3.2Homogeneous potentials

12.3.3Higher variational equations

12.4Application to parametrized potentials

12.4.1Space of germs of integrable potentials

12.4.2Eigenvalue bounding of some n-body problems

Jacques Féjoz

Introduction to KAM theory with a view to celestial mechanics

13.1Twisted conjugacy normal form

13.2One step of the Newton algorithm

13.3Inverse function theorem

13.4Local uniqueness and regularity of the normal form

13.5Conditional conjugacy

13.6Invariant torus with prescribed frequency

13.7Invariant tori with unprescribed frequencies

13.8Symmetries

13.9Lower dimensional tori

13.10Example in the spatial three-body problem

AIsotropy of invariant tori

BTwo basic estimates

CInterpolation of spaces of analytic functions

Marek Grochowski and Wojciech Kryński

Invariants of contact sub-pseudo-Riemannian structures and EinsteinWeyl geometry

14.1Introduction

14.2Dimension 3

14.3EinsteinWeyl geometry

14.4Dimension 2n + 1

14.5Contact sub-pseudo-Riemannian symmetries

14.6Appendix: Isometries in dimension 5

Jérôme Lohéac and Jean-François Scheid

Time-optimal control for a perturbed Brockett integrator

15.1Introduction

15.2Controllability and time-optimal controllability

15.3An approximate linearized time-optimal control problem

15.4Numerical computation of a time-optimal trajectory

15.4.1Finite-dimensional minimization problem

15.4.2Numerical example

15.5Application to time-optimal micro-swimmers

15.5.1Modeling and problem formulation

15.5.2Numerical computation of a time-optimal trajectory

15.6Conclusion

Jean-Pierre Marco

Twist maps and Arnold diffusion for diffeomorphisms

16.1From Arnold diffusion to twist maps

16.2Setting and main result

16.3Proof of Theorem 1

16.3.1Choice of ε > 0

16.3.2Implementation of Moeckels method

16.3.3Normally hyperbolic shadowing

16.3.4Conclusion of the proof of Theorem 1

Gianna Stefani and Pierluigi Zezza

A Hamiltonian approach to sufficiency in optimal control with minimal regularity conditions: Part I

17.1Introduction

17.1.1The problem

17.2Notations and preliminary results

17.2.1Symplectic notations

17.2.2The Pontryagin maximum principle

17.3A Hamiltonian approach to optimality

17.3.1The Cartan form

17.3.2The super-Hamiltonian and its properties

17.3.3Abstract sufficient optimality conditions

17.3.4The minimum time problem

17.4Final comments

Index

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.144.255.55