Spatial Coherence for Visual Motion Analysis

1. Front Matter

   PDF

2. 2D Motion Description and Contextual Motion Analysis: Issues and New Models

   • P. Bouthemy

   Pages 1-15

3. Structure from Periodic Motion

   • Serge Belongie, Josh Wills

   Pages 16-24

4. 3D SSD Tracking from Uncalibrated Video

   • Dana Cobzas, Martin Jagersand

   Pages 25-37

5. Comparison of Edge-Driven Algorithms for Model-Based Motion Estimation

   • Hendrik Dahlkamp, Artur Ottlik, Hans-Hellmut Nagel

   Pages 38-50

6. On the Relationship Between Image and Motion Segmentation

   • Adrian Barbu, Song Chun Zhu

   Pages 51-63

7. Motion Detection Using Wavelet Analysis and Hierarchical Markov Models

   • Cédric Demonceaux, Djemâa Kachi-Akkouche

   Pages 64-75

8. Segregation of Moving Objects Using Elastic Matching

   • Vishal Jain, Benjamin B. Kimia, Joseph L. Mundy

   Pages 76-90

9. Local Descriptors for Spatio-temporal Recognition

   • Ivan Laptev, Tony Lindeberg

   Pages 91-103

10. A Generative Model of Dense Optical Flow in Layers

    • Anitha Kannan, Brendan Frey, Nebojsa Jojic

    Pages 104-114

11. Analysis and Interpretation of Multiple Motions Through Surface Saliency

    • Mircea Nicolescu, Changki Min, Gérard Medioni

    Pages 115-126

12. Dense Optic Flow with a Bayesian Occlusion Model

    • Kevin Koeser, Christian Perwass, Gerald Sommer

    Pages 127-139

13. Back Matter

    PDF

Motionanalysisisacentralproblemincomputervision,andthepasttwodecades have seen important advances in this ?eld. However, visual motion is still often considered on a pixel-by-pixel basis, even though this ignores the fact that image regions corresponding to a single object usually undergo motion that is highly correlated. Further, it is often of interest to accurately measure the boundaries of moving regions. In the case of articulated motion, especially human motion, discovering motion boundaries is non-trivial but an important task nonetheless. Another related problem is identifying and grouping multiple disconnected - gions moving with similar motions, such as a ?ock of geese. Early approaches focused on measuring motion of either the boundaries or the interior, but s- dom both in unison. For several years now, attempts have been made to include spatial coherence terms into algorithms for 2- and 3-D motion recovery, as well as motion boundary estimation. This volume is a record of papers presented at the First International Wo- shop on Spatial Coherence for Visual Motion Analysis, held May 15th, 2004 in Prague, in conjunction with the European Conference on Computer Vision (LNCS 3021–4). The workshop examined techniques for integrating spatial - herence constraints during motion analysis of image sequences. The papers were revised after the workshop to allow for incorporation of feedback from the workshop.

• 3D
• 3D imaging
• algorithm
• algorithms
• classification
• cognition
• computer vision
• image analysis
• image sequence
• motion detection
• motion estimation
• multiple motion detection
• neural networks
• segmentation
• spatio-temporal recognition
• algorithm analysis and problem complexity

• University of Toronto, Toronto, Canada

  W. James MacLean

Policies and ethics