压缩传感在无线视频监控中的应用研究*

图像采集数据量大是制约视频监控系统向无线化方向发展的主要因素，提出利用压缩传感进行视频图像的采样，为无线视频监控带来一种新的应用研究。为了减少图像稀疏分解过程的计算量和存储量，在匹配追踪算法的基础上，引入量子遗传算法，实现快速的图像稀疏表示。以Fourier矩阵作为压缩传感的测量矩阵，能有效减少测量数据量，并提高重构图像的质量。仿真实验证明，采用压缩传感所得到的测量数据量远小于传统采样方法所获的数据量，突破了传统信号采样的瓶颈，提高了采样效率，最终获取的压缩测量值能够很好地恢复为监控场景。

Wireless image and video surveillance based on compressive sensing

Abundant data of image capture is the main factor that hinders the video surveillance system to develop in the direction of wireless. This paper introduced compressed sensing into video image sampling, bringing a new kind of application research for wireless video surveillance. In order to reduce the computation and storage capacity in sparse decomposition process, based on the matching pursuit algorithm, introduced quantum genetic algorithm to implement fast sparse representation of images. And regarded fourier matrix as the measurement matrix of compressed sensing, it could reduce the amount of measurement data effectively, improve the quality of reconstructed image. The simulation results show that the amount of measurement data received by compressed sensing is far less than that obtained by traditional sampling method, breaking the bottleneck of traditional signal sampling, improving sampling efficiency, and the compression measurements eventually acquire can well restore the surveillance scene.