Abstract

Block compressed sensing of video based on unstable sampling rates and multihypothesis predictions

Author(s): Zhao Chenglin, Liu Jiangyun

Existing divide block video compression perception is usually used the same measurement matrix measure for all image block, the average distribution of sampling ratemeasurementmethod ignores at different areas of the structure complexity and change degree different of the facts in the video. In order to solve this problem, this paper proposed an adaptive allocation sampling rate of the unstable sampling rate compressionmethod by the distribution characteristics of the video inter-frame correlation. Categorizing the size of the image block according to the inter-frame correlation and assign different sampling rates, refactoring process adopt the unstable sampling rate multihypothesis predictions algorithm to make full use of the inter-frame correlation. Experimental results show that this paperÂ’s algorithmcan reconstruct the high quality video images under low sampling rate, and the unstable sampling ratemeasurementmethod is helpful to improve the refactoring quality of sports areas