Title :
Hierarchical threshold technique oriented to video object segmentation
Author :
Zheng, JiaLi ; Zhang, Yongdong ; Ni, GuangNan
Author_Institution :
Graduate Sch. of the Chinese Acad. of Sci., Beijing
fDate :
Nov. 28 2007-Dec. 1 2007
Abstract :
This paper proposed a motion vector error function to segment the objects in video sequence. The key point of the segmentation technique is how to set the appropriate threshold to distinguish the global motion region from the local motion region precisely. This paper also proposed a hierarchical threshold technique based on global motion estimation to solve this problem. Experimental results show that the proposed techniques are robust techniques which refine the set of global motion pixels hierarchically and segment the video objects effectively.
Keywords :
image motion analysis; image resolution; image segmentation; image sequences; video signal processing; global motion pixels; global motion region; hierarchical threshold technique; motion vector error function; video object segmentation; video sequence; Cameras; Image segmentation; Information processing; Laboratories; Layout; Motion estimation; Object segmentation; Parameter estimation; Pixel; Video signal processing; Global motion estimation; threshold method; video object segmentation;
Conference_Titel :
Intelligent Signal Processing and Communication Systems, 2007. ISPACS 2007. International Symposium on
Conference_Location :
Xiamen
Print_ISBN :
978-1-4244-1447-5
Electronic_ISBN :
978-1-4244-1447-5
DOI :
10.1109/ISPACS.2007.4446023