DocumentCode
3499989
Title
Shape consistent segmentation algorithm for extraction of moving object
Author
Kyeong Joong Kim ; Suh, Jong Yeui ; Lee, Doo Hyun ; Lim, Chae Wook ; Park, Kyu Tae
Author_Institution
Dept. of Electron. Eng., Yonsei Univ., Seoul, South Korea
Volume
2
fYear
1996
fDate
14-18 Oct 1996
Firstpage
902
Abstract
We propose a new segmentation algorithm for object extraction as a part of a region based coding scheme. The segmentation is based on spatial and temporal information with emphasis on the shape consistency of the region. The segmentation algorithm presented here utilizes splitting of the region into smaller regions based on edge information, followed by merging using motion similarity and shape consistency. The results show that the proposed method yields objects with low model failure rates
Keywords
edge detection; feature extraction; image coding; image segmentation; edge information; low model failure rates; motion similarity; moving object extraction; region based coding scheme; shape consistency; shape consistent segmentation algorithm; Data mining; Degradation; Image coding; Image edge detection; Image segmentation; Image sequences; Morphological operations; Motion estimation; Pixel; Shape control;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing, 1996., 3rd International Conference on
Conference_Location
Beijing
Print_ISBN
0-7803-2912-0
Type
conf
DOI
10.1109/ICSIGP.1996.566234
Filename
566234
Link To Document