Quality improvement through an efficient image segmentation

The next generation coding techniques are closely connected to the major techniques on computer vision and computer graphics especially in the field of segmentation. This paper deals with a study on the foreground and background segmentation in videophone-type sequence and tries to use it for quality enhancement of reconstructed image. A simple 1-D high-pass mask is used to reduce redundant correlated image information. And an edge information from edge operator is incorporated to help to find the contour of object boundary. An appropriate bridging strategy is proposed as one part of constructing the linear contour. Simulation results show that the proposed algorithm works well for videophone-type sequence and improves the reconstructed image quality considerably.