Research On The Image Adaptive Coding And Reconstruction Algorithm Based On Compressed Sensing

With the rapid development of acquisition,display and processing technology of digital media,new applications and services of various high quality images and videos are emerging,which makes the image/video data exploding.Huge amounts of image/video data put forward high requirements to transport and storage.How to realize the efficient compression has become the long-standing challenge in the field of image and video coding and decoding.In the recent years,the emerging compressive sensing theory greatly improves the compression ratio of the signal and reduces the pressure of storage and transmission in the signal processing.It is undoubtedly a great innovation and progress for the research in the field of image and video decoding.This paper briefly introduces the existing image coding algorithm based on compressive sensing,and emphatically introduces block compressive sensing theory,sparsity determination criterion and the adaptive image coding algorithm using the spatial correlation of the image.On this basis,the paper proposes an image adaptive encoding algorithm and two kinds of improved reconstruction algorithms for image sequence.The specific contents are as follows:(1)An image adaptive encoding algorithm based on compressed sensing: In the case of meeting the requirements of sampling rate on the encoding side,different sampling rates can be reasonably allocated for every block in the image according to the sparsity of the image block in the TV domain,so the compression ration of the image can be improved and the high quality reconstructed images can be got.(2)An time domain enhancement algorithm based on adaptive kalman: On the basis of MC-BCS-SPL algorithm,the noise distribution characteristics of the image sequence are analyzed in this paper.Then the adaptive kalman filtering is applied to improve the effect of time-domain reconstructed images.In this case,it removes the noise of the inter-frame effectively and improves the subjective effect of the image greatly.(3)A reconstruction algorithm of image sequence based on TVAL3: In this paper,a reconstruction algorithm which combines TVAL3 and the new three step search method is proposed to reconstruct the image sequence.In this algorithm,the TVAL3 algorithm is used as the reconstructed algorithm of the image,and the new three step search method(NTSS)is used as block match algorithm in order to obtain the optimal matching block in reference frame of current frame.After the image sequence is reconstructed using the above method,the wiener filter is used to get better subjective images.The experimental results show that the proposed image adaptive encoding algorithm based on compressed sensing reduces the sampling data of the encoding side and achieves a high efficiency of compression.Simultaneously,two kinds of improved reconstruction algorithms for image sequence can also improve the quality of the reconstruction image sequence effectively.