Loading...
Please wait, while we are loading the content...
Similar Documents
Joint disparity motion estimation and segmentation for object oriented stereoscopic image coding
| Content Provider | Semantic Scholar |
|---|---|
| Author | Yang, Cheng Hong Konrad, Janusz |
| Copyright Year | 2007 |
| Abstract | Rapport technique de l'INRS-T el ecommunications no. 97-05 This report describes the Ph.D thesis proposal of Cheng Hong Yang prepared under the supervision of Prof. Janusz Konrad and defended on April 18 1997 in front of a committee consisting of Prof. Summary This document describes a project aiming at the development of a new approach to video segmentation. In particular, in order to divide individual video frames into regions corresponding to objects from a 3-D scene, the proposed approach attempts to jointly exploit multiple sources of visual information, such as motion, disparity and luminance/color. The resulting segmentation is expected to nd applications in object-oriented coding of monoscopic and stereoscopic video sequences. The proposed approach is based on Markov random eld models linked together by the maximum a posteriori probability estimation criterion. In the formulation robust error functions are used to minimize the impact of outliers, while the optimization stage is based on a continuation method. Encouraging initial experimental results for the case of motion segmentation are demonstrated. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.bu.edu/vip/files/pubs/reports/Yang97-05inrs.pdf |
| Alternate Webpage(s) | http://www.inrs-telecom.uquebec.ca/users/viscom/english/research/local/reports/Yang97-05inrs.ps.gz |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |