Variational, Geometric, and Level Set Methods in Computer Vision

1. Front Matter

   PDF

2. A Study of Non-smooth Convex Flow Decomposition

   • Jing Yuan, Christoph Schnörr, Gabriele Steidl, Florian Becker

   Pages 1-12

3. Denoising Tensors via Lie Group Flows

   • Y. Gur, N. Sochen

   Pages 13-24

4. Nonlinear Inverse Scale Space Methods for Image Restoration

   • Martin Burger, Stanley Osher, Jinjun Xu, Guy Gilboa

   Pages 25-36

5. Towards PDE-Based Image Compression

   • Irena Galić, Joachim Weickert, Martin Welk, Andrés Bruhn, Alexander Belyaev, Hans-Peter Seidel

   Pages 37-48

6. Color Image Deblurring with Impulsive Noise

   • Leah Bar, Alexander Brook, Nir Sochen, Nahum Kiryati

   Pages 49-60

7. Using an Oriented PDE to Repair Image Textures

   • Yan Niu, Tim Poston

   Pages 61-72

8. Image Cartoon-Texture Decomposition and Feature Selection Using the Total Variation Regularized L ¹ Functional

   • Wotao Yin, Donald Goldfarb, Stanley Osher

   Pages 73-84

9. Structure-Texture Decomposition by a TV-Gabor Model

   • Jean-François Aujol, Guy Gilboa, Tony Chan, Stanley Osher

   Pages 85-96

10. From Inpainting to Active Contours

    • François Lauze, Mads Nielsen

    Pages 97-108

11. Sobolev Active Contours

    • Ganesh Sundaramoorthi, Anthony Yezzi, Andrea Mennucci

    Pages 109-120

12. Advances in Variational Image Segmentation Using AM-FM Models: Regularized Demodulation and Probabilistic Cue Integration

    • Georgios Evangelopoulos, Iasonas Kokkinos, Petros Maragos

    Pages 121-136

13. Entropy Controlled Gauss-Markov Random Measure Field Models for Early Vision

    • Mariano Rivera, Omar Ocegueda, Jose L. Marroquin

    Pages 137-148

14. Global Minimization of the Active Contour Model with TV-Inpainting and Two-Phase Denoising

    • Shingyu Leung, Stanley Osher

    Pages 149-160

15. Combined Geometric-Texture Image Classification

    • Jean-François Aujol, Tony Chan

    Pages 161-172

16. Heuristically Driven Front Propagation for Geodesic Paths Extraction

    • Gabriel Peyré, Laurent Cohen

    Pages 173-185

17. Trimap Segmentation for Fast and User-Friendly Alpha Matting

    • Olivier Juan, Renaud Keriven

    Pages 186-197

18. Uncertainty-Driven Non-parametric Knowledge-Based Segmentation: The Corpus Callosum Case

    • Maxime Taron, Nikos Paragios, Marie-Pierre Jolly

    Pages 198-209

19. Dynamical Statistical Shape Priors for Level Set Based Sequence Segmentation

    • Daniel Cremers, Gareth Funka-Lea

    Pages 210-221

20. Non-rigid Shape Comparison of Implicitly-Defined Curves

    • Sheshadri R. Thiruvenkadam, David Groisser, Yunmei Chen

    Pages 222-234

Mathematical methods has been a dominant research path in computational vision leading to a number of areas like ?ltering, segmentation, motion analysis and stereo reconstruction. Within such a branch visual perception tasks can either be addressed through the introduction of application-driven geometric ?ows or through the minimization of problem-driven cost functions where their lowest potential corresponds to image understanding. The 3rd IEEE Workshop on Variational, Geometric and Level Set Methods focused on these novel mathematical techniques and their applications to c- puter vision problems. To this end, from a substantial number of submissions, 30 high-quality papers were selected after a fully blind review process covering a large spectrum of computer-aided visual understanding of the environment. The papers are organized into four thematic areas: (i) Image Filtering and Reconstruction, (ii) Segmentation and Grouping, (iii) Registration and Motion Analysis and (iiii) 3D and Reconstruction. In the ?rst area solutions to image enhancement, inpainting and compression are presented, while more advanced applications like model-free and model-based segmentation are presented in the segmentation area. Registration of curves and images as well as multi-frame segmentation and tracking are part of the motion understanding track, while - troducing computationalprocessesinmanifolds,shapefromshading,calibration and stereo reconstruction are part of the 3D track. We hope that the material presented in the proceedings exceeds your exp- tations and will in?uence your research directions in the future. We would like to acknowledge the support of the Imaging and Visualization Department of Siemens Corporate Research for sponsoring the Best Student Paper Award.

• 3D modeling
• 3D vision
• Active contour
• Computer Vision
• Image segmentation
• Mapping
• Shading
• Stereo
• Textur
• classifier systems
• image analysis
• image enhancement
• motion analysis
• stereo vision
• topology
• algorithm analysis and problem complexity

• MAS - Ecole Centrale Paris, Grande Voie des Vignes, Chatenay-Malabry, France

  Nikos Paragios

• I.N.R.I.A, 2004 route des lucioles,, Sophia-Antipolis, France

  Olivier Faugeras

• Department of Mathematics, UCLA,  

  Tony Chan

• Image and Pattern Analysis Group, Heidelberg Collaboratory for Image Processing, University of Heidelberg, Germany

  Christoph Schnörr

Policies and ethics