Title :
A new mesh based temporal-spatial segmentation for image sequence
Author :
Kuo, Chung-Ming ; Hsieh, Chaur-Heh ; Huang, Yong-Ren ; Sun-Lon, Zen
Author_Institution :
Dept. of Electr. Eng., Chinese Naval Acad., Kaohsiung, Taiwan
Abstract :
Video segmentation, which identifies an object from an image sequence, is a crucial issue for object-based video processing. This paper presents a mesh-based silhouette segmentation technique for video sequences. We first define an active motion activity as a temporal measure, and then a new maximum a-posteriori (MAP) scheme for spatial segmentation is developed. We also introduce clique potentials for the Gibbs distribution such that the edge connectivity and the smoothness of the homogenous patches are guaranteed. Simulation results indicate that the new method achieves very good segmentation results with a much lower computational cost, as compared to existing techniques
Keywords :
image segmentation; image sequences; maximum likelihood sequence estimation; mesh generation; video signal processing; Gibbs distribution; active motion activity; clique potentials; computational cost; edge connectivity; homogenous patch smoothness; image sequences; maximum a-posteriori scheme; mesh-based silhouette segmentation technique; mesh-based temporal-spatial segmentation; object identification; object-based video processing; simulation; temporal measure; video objects; video segmentation; video sequences; Computational efficiency; Computational modeling; Image coding; Image segmentation; Image sequences; Motion measurement; Size measurement; Video coding; Video compression; Video sequences;
Conference_Titel :
Computer Software and Applications Conference, 2000. COMPSAC 2000. The 24th Annual International
Conference_Location :
Taipei
Print_ISBN :
0-7695-0792-1
DOI :
10.1109/CMPSAC.2000.884756