Font Size: a A A

Image Denoising Method Based On The Low Rank Texture

Posted on:2019-06-24Degree:MasterType:Thesis
Country:ChinaCandidate:L Z LiFull Text:PDF
GTID:2428330548992630Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Image is the primary source for people to obtain information,and as a major carrier of visual information,image is an important tool for human to obtain and use information.However,at various stages,such as generation,transmission and feedback,image can be disturbed by noise,which brings many complicated problems to the subsequent image processing.Therefore,image denoising has become an important research direction in image restoration at the present stage.In this paper,based on the in-depth study of the theory of compressed sensing and threshold truncation,a de-noising model based on iterative logarithmic threshold is proposed.The main work of this article is as follows:On the basis of the traditional soft threshold algorithm,a sparse reconstruction based on the iterative soft threshold algorithm is introduced on the basis of the compression perception theory.For the soft threshold truncation algorithm cannot achieve both reduce the noise and retaining the details of the problem,introduced the logarithm threshold algorithm,compared with the soft threshold algorithm,log truncation algorithm of threshold threshold is more reasonable and smooth it is just steady alternating between soft and hard threshold decreased,the reconstruction result is more smooth stable.In the low rank matrix recovery based on the first frame denoising in non local image by matching the similarity block using linear correlation between image blocks,low rank image block and the composition of the corresponding matrix,then the matrix block decomposition of singular value decomposition of the logarithmic threshold truncation,by iteration,the non noise approximation,finally reducing image.
Keywords/Search Tags:Image de-noising, logarithmic threshold, low rank approximation, non-local self-similarity, compression sensing, singular value decomposition
PDF Full Text Request
Related items