Research Of Distributed Video Codec And Image Super-resolution Based On Compressed Sensing

Sampling is the bridge between analog source signal and digital signal. With the rapid progress of information technologies, the demands for information are increasing dramatically. So the existing systems are very difficult to meet the challenges of high speed sampling, large volume data transmission and storage. How to acquire information in signal efficiently is an urgent problem in electronic information fields. In recent years, an emerging theory of signal acquirement--compressed sensing (CS) provides a golden opportunity for solving this problem. This paper surveys the principles of compressed sensing and its related applications. Some further works on this theory are also presented. In this paper, an intelligent video analysis system for distance learning is proposed, which apply the intelligent video analysis algorithms in the video shooting.For a lower complexity of distributed coding requirements, a new video codec without using feedback channel which combines compressed sensing theory and some critical ideas of traditional video codecs is proposed. The encoder works as a projector, which projects a high dimensional signal to a low dimensional domain, while the decoder which is no longer the inverse of encoder works for the solution of underdetemiined equations by constructing joint sparse representation (JMS). Compared with traditional codecs, the new codec is simpler. Not only samples needed in the new codec are fewer than in traditional codecs and the number of samples can be changed according to the encoding mode, but also rate-distortion (RD) optimization through adaptive measurement collection are obtained. The experimental results demonstrate the effectiveness of the proposed method.Based on CS, we introduce sparse signal representation theory to modify the local geometric similarity model and construct sparse geometric similarity representation. Based on the modified model we can estimate the optimized reconstruct coefficients by jointing the original global and local image structure themselves. What is more, proposed K-SVD Algorithm helps to update dictionary without the support of other training image database. The experimental results show that the algorithm can greatly improve the reconstruction of the edge and texture details in the high-resolution image.