An improved framework of distributed video codec

Recently, distributed video coding (DVC) has become a hot research topic, as its low complexity encoder and high complexity decoder can suit for many emerging applications. However, the practical performance gap between DVC and traditional video coding is still large and need more research about it. In this paper, we propose to exploit the capabilities of side information (SI) extraction, online estimation of correlation noise model (CNM) parameters and joint bitplane decoding to construct an integrated and improved DVC framework for better system performance. The SI extraction along with online CNM parameters estimation is exploited to help decoding, and the joint bit plane decoding that we propose in this paper is used to reduce the transmission rate for the same PSNR performance. The experimental results show that the proposed DVC architecture could work efficiently and improve the PSNR performance by 2-3 db for the test sequence.