Segmentation Algorithm by Multicriteria Region Merging
作者： B. Marcotegui
DOI: 10.1007/978-1-4613-0469-2_36
关键词:
摘要: This paper presents a segmentation algorithm for image sequences, by multicriteria region merging. The output of connected filter is simplified iteratively merging the two most similar adjacent regions, while given representation quality preserved. We have defined several region-similitude criteria: grey-level, texture and motion resemblance. Texture criteria introduce feedback “segmentation-coding step” that improves coding efficiency.
springer.com
sci-hub.st 参考文章(4)
Olivier Monga, Segmentation d'images par croissance hiérarchique de régions Paris 11. ,(1988)
Jean C. Serra, Philippe Salembier, Connected operators and pyramids SPIE's 1993 International Symposium on Optics, Imaging, and Instrumentation. ,vol. 2030, pp. 65- 76 ,(1993) , 10.1117/12.146672
Montse Pardàs Feliu, Philippe Jean Salembier Clairon, Time-recursive segmentation of image sequences M.J.J. HOLT. EURASIP.. pp. 18- 21 ,(1994)