Object segmentation based on multiple features for low bit rate video coding

Video object segmentation for object based video coding according to MPEG-4 should be able to combine many features of the image sequence to achieve high quality object segmentation. The significant features include color, motion vector and change information. This paper presents an object segmentation algorithm where the image features are combined for use in the segmentation process. A fast recursive shortest spanning tree (FRSST) algorithm is used to segment the image into different intensity regions because of the good image segmentation performance and the fast algorithm. Motion vectors are estimated from consecutive frames and searched by thresholding hierarchical block matching (HBM) which require low computation and give acceptable results. After that many features are integrated in the segmentation decision process. Then, the postprocessing refines the final segmentation output. The results from some test sequences especially in a video conference sequence have a good performance and show the object of interest boundary clearly