Title :
Research of GrabCut algorithm for single camera video synopsis
Author :
Lili Hao ; Jianrong Cao ; Chengdong Li
Author_Institution :
Shandong Provincial Key Lab. of Intell. Building Technol., Shandong Jianzhu Univ., Jinan, China
Abstract :
In this paper, the GrabCut segmentation algorithm is used for the moving object matting in single camera surveillance video. In the case of the general location of the moving object has been drawn, we can get more accurate segmentation result and the process of user interaction is simplified. At the same time, the segmentation results are placed into one background image to concentrate the surveillance video in time and space and a single camera surveillance video synopsis can be formed. The implementation result shows that video synopsis is an effective way to compress the video and the browsing efficiency is improved in large extent.
Keywords :
cameras; image motion analysis; image segmentation; user interfaces; video coding; video surveillance; GrabCut algorithm; GrabCut segmentation algorithm; accurate segmentation; background image; browsing efficiency; camera surveillance video synopsis; moving object matting; single camera surveillance video; single camera video synopsis; user interaction; video compression; Algorithm design and analysis; Cameras; Educational institutions; Image color analysis; Image segmentation; Minimization; Surveillance;
Conference_Titel :
Intelligent Control and Information Processing (ICICIP), 2013 Fourth International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4673-6248-1
DOI :
10.1109/ICICIP.2013.6568151
