Motion Compensation by Affine Transform using Edge Infomation of Image

A BMA(Block Matching Algorithm) causes block artifacts in very low bit rate video coding, accordingly, a motion compensation method using an transform has been proposed as a solution to this problem. When image are divided using an affine transform, there is an increase in the prediction errors of the compensated image and distortion will occur as image object do not necessarily conform to triangular patterns. In this paper, we propose a method that first divides an image into triangular patches according to its edge information and than divides the image into more detailed triangular patches where a higher standard deviation of difference image occurs. The experimental results indicate that the proposed scheme improve both the objective and the subjective quality of an image when compared with H.263