会议专题

Moving Objects Segmentation in Video Sequences

Moving objects segmentation accurately is one of the most challenging problems in computer vision.This paper present a novel approach to segment moving obiects with edge information and temporal information using graph cuts model.To improve this problem,motion segmentation problem is considered as an image-labeling process in which every pixel in the image will be assigned a labeI describing its motion between consecutive frames.And a graph cuts based energy minimization method is used by define a measure function of smoothness and accuracy of motion vectors to achieve this goal.And to improve the accurate of the motion edge map,an improved ME(Moving Edge)map acquiring method.is proposed.Experiments and results show that oar algorithm is efficient and noise robot.

graph cut moving object segmentation

Kuang Ping Jianping Li Gu XiaoFeng Liao Jianming

国际会议

The International Conference Information Computing and Automation(2007国际信息计算与自动化会议)

成都

英文

382-384

2007-12-19(万方平台首次上网日期,不代表论文的发表时间)