A New Architecture of Distributed Video Coding For Real Time Systems

Most of the reported Distributed Video Coding (DVC) schemes have a high time-delay in decoder which hinders its practical application in real-time systems. In this paper, we propose a novel DVC architecture based on MPI cluster in a practical scene, which includes three major approaches to accelerate decoder: an efficient parallel algorithm for generation of the Side Information (SI), a parallel method for LDPC iterative decoding process and a rate control approach based on machine learning for reducing feedbacks from decoder to encoder. The experimental results on cluster demonstrate significant savings in time with a little loss of video quality.