A new approach in the object-oriented coding at low bit rates

The object-oriented coding technique segments each image of a video sequence into a number of moving objects and compensates the motion of each object. Thus it can reconstruct real motion better than the block-based coding algorithm at low bit rates. It uses mapping parameters to estimate motion information of each object. To obtain the good quality of a reconstructed image at a low bit rate, we introduce a new algorithm to the object-oriented coding algorithm by using variable length coding technique for the foreground regions after detecting the changed regions. In addition, the border information obtained from our segmentation technique is used to compress the first frame of the video sequence. Computer simulation results with a test sequence show that the proposed method achieves better results than the H.261 standard both in the quantitative measurements at 64 kbits/sec.