Spatial-motion clustering for video compression

摘要

The most significant contribution of this paper is the algorithm for the segmentation of videoconference image sequences for content-based coding. The segmentation is based on motion estimation trough the computation of the optical flow field and motion segmentation by applying a graph-theoretical clustering. A visual collaboration environment for desktop is being implemented and tested. The segmentation algorithm is added to the video codec of this videoconference environment which is based on the OpenH323 project. In addition, an overview of the state of the art of segmentation based video coding aimed at low-bit rate application is presented.