An Overview of Video Object Segmentation

With a rapid demand for multimedia information, visual information plays an important part in information achieving. But the data size of a video is so enormous that it should be compressed and coded immediately. In order to improve the efficiency of MPEG-4 and MPEG-7 adopt content-based coding method, a video sequence is segmented into several meaningful video objects before coding. Video object segmentation plays a critical role in realizing content-based coding and description, which has an immediate influence on coding efficiency and description validity. Furthermore, the video object segmentation technique is also widely used in our real life, such as video monitoring, video conference and some other areas. So many segmentation algorithms have been proposed, but the development of video object segmentation is still in its primary stage. Some of the disadvantages should be improved in the future.