Research Of Distributed Video Code And Transmission Scheme Based On Compressed Sensing In Wireless Multimedia Sensor Networks

Compressed sensing breaks Nyquist–Shannon sampling theorem, reduces the sampling rate greatly by combining compression with sampling. Distributed compressive sensing transfers computation complexity from the encode side to the decode side. This feature meets the requirements of low complexity and low computation in wireless multimedia sensor network. With the rapid development of wireless multimedia sensor network, distributed compressive video sensing becomes research hotspots in academia and Engineering.This paper introduces the basic theoretical knowledge of compressed sensing and some research hotspots, including sparse representation, measurement matrix and reconstruction algorithms. This paper focuses on the frame of distributed compressive sensing and optimized transmission of compressed sensing video stream in wireless channel. This paper studies the effect of quantize matrix to reconstruction quality and creates more suitable quantize matrix to compressed sensing. The main contents and innovation points are as follows:(1) This paper analyses distribution features of compressed sensing image coefficient and proposes adaptive quantize matrix. Compared with uniform quantization and H.264 standard quantization, adaptive quantize matrix leads to higher reconstruction quality. Adaptive quantize matrix has the same performance with Lloyd quantize matrix, but the computation complexity of Lloyd quantize matrix is much higher than adaptive quantize matrix.(2)This paper researches the effect of channel noises on compressed sensing reconstruction quality and proposes dynamic rate control scheme. This scheme adjusts measure rate of each block to counterpoise the reconstruction quality.This scheme also raises sampling rate when reconstruction quality is affected by channel noise to obtain higher reconstruction quality.