Measurement compression in distributed compressive video sensing

In some application scenarios a video codec with simple encoder and complex decoder is desired. Distributed video coding (DVC) and compressive sensing (CS) theory proposed recently are two techniques suitable to such scenarios, and several video coding schemes that combine CS with DVC have appeared. However, in these existing integrated schemes, the dependencies between measurements of successive image frames have not yet been exploited. In this paper, we first propose a Gaussian distribution model to describe the correlations between the measurements of a CS frame and the measurements of its side information. And then, we propose to compress the measurements of CS frame using a channel coder similar to that in DVC. Experiment results indicate that the compression ratio of the proposed measurement compression scheme achieves 56.50%–71.66% for four test sequences when measurement rates for key frame and CS frame are 50% and 20% respectively.