Stereo Image Sequence Coding Based on Phase Correlation Disparity Estimation

This paper will discuss how to compress the stereo media data in order to avoid getting poor performance when transmitting these data over network. We proposed a novel disparity estimation method for stereo video compression. Based on our method, we overcome some of the defects of phase correlation and improve disparity estimation accuracy on stereo image. From our experiments, the results not only show higher compression performance but also better than H.264 compression standard. In computing accelerated testing experiment, our proposed method have higher disparity estimation efficiency than block matching, especially when the resolution of the video are growing up or the foreground in a single frame occupies the most area of the scene.