Font Size: a A A

Segmentation Of MPEG-4 Video Object Plane Using Spatial And Temporal Information

Posted on:2005-12-09Degree:MasterType:Thesis
Country:ChinaCandidate:F WangFull Text:PDF
GTID:2168360122487683Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the development of multimedia technique, the need of people for multimedia information is translating from simple playing to accessing searching and manipulating based on content, and transport stream should be suit to different networks and all kind of terminals. So MPEG-4 is an ISO/IEC standard being developed by MPEG(Moving Picture Expert Group), which adopt coding scheme based on object and is asked to analysis images and video sequences. Coding scheme based on object is an important characteristic, but object segmentation is not completed satisfactorily yet. In order to apply the MPEG-4 visual standard, each frame of video sequence should be represented in terms of video object planes(VOPs). Video objects in still pictures or video sequences should be first identified before the encoding process starts. It requires a prior decomposition of sequences into VOPs so that each VOP represents an object.MPEG-4 standard and the development of image segmentation is introduced in this paper. The automatic segmentation using spatial and temporal information is proposed in this paper. The continuity of time and space is used sufficiently. The basic idea is to use the difference of adjacent frames, texture information and color information of current frame, face localization technology to initialize object, then to employ the quick Hausdorff distance method to track object. The method and process which is used is discussed particularly. The proceeding result shows good segmentation by the approach.
Keywords/Search Tags:MPEG-4, image segmentation, canny, morphology
PDF Full Text Request
Related items