Research On Image De-noising Theory Based On Compressed Sensing

The core of compressive sensing theory is to reconstruct the signal accurately and recover the original signal with the less sampling data.. This theory is a new star in image processing, and it has been shown another way of dealing with the image.. It also caused the researchers’ strong interest. In this paper, The research is how to combine the compressed sensing theory with the image denoising, so as to remove the image noise as well as possible, and not make the image blur.In the process of image transmission and compression, it will inevitably be affected by the noise, so it will lead to the quality of the image degradation, which is what we want to see. So image denoising becomes a hot research topic in the field of digital image processing.. As the research is deep, there are new denoising methodsThis paper takes the development of compressed sensing as the background. The to compress the perception of the development of history as the background, combined with domestic and foreign research results, has carried on the detailed narration, especially the compressed sensing constitute the three pillars: sparse signal said, a detailed analysis of the design of measurement matrix and signal reconstruction algorithms. This paper is focus on p=1(1 norm) and 0<p<1 two cases(combining the values of the two p). The fast reconstruction algorithm of p=1 norm gradient projection based on TV(1 norm) is studied and based on the image denoising algorithm TV norm algorithm, an algorithmpTV RLS for the case of p ?1 2(1 2 norm) is presented, And the feasibility of the algorithm is verified by simulation experiments, which can preserve the effective information of the image while better denoising. In this paper, the theory of the compressed sensing theory is applied to the pavement crack image denoising, which lays the foundation for the further treatment of the pavement cracks.