Cartoon Texture Image Decomposition And Total Variation Regularization For Image Restoration

Image restoration technology is one of the important research projects in the field of image processing. Image restoration is to use some prior knowledge of the image degradation phenomenon and restoration algorithm to rebuild the original image, in order to improve image clarity, image fidelity, and eliminate noise. This paper proposes a kind of approach by using cartoon texture image decomposition and a generalized accelerated proximal gradient algorithm to realize image restoration.In this paper, first study the classic image degradation model and the total variation regularization model, and to promote accelerated proximal gradient algorithm for generalized accelerated proximal gradient algorithm to solve the total variation model, so as to improve the convergence speed and peak signal to noise ratio. Because of the sensitive degree of high frequency and low frequency component of noise image is different, so we can separate the high frequency and low frequency component from blurred image, and then choose respectively different regularization parameter for the high frequency and low frequency components to restore, and then to compose weight synthesis.In this paper, for flexible separation of high frequency, low frequency component of image, and put forward the texture scale decomposition of parameters are adjustable to meet different requirements of cartoon texture decomposition algorithm. Finally, tries to explore how to realize the regularization parameter and the synthetic weighted coefficient of the optimal selection, adaptive choices.In this paper, use MATLAB to do the simulation experiment. First, to decompose the original fuzzy image into cartoon part and texture part, cartoon part is the low frequency component of the image, noise interference is small, the texture of the image is the high frequency component of the image, noise interference is sharp, and then to de-blur and de-noise by using the generalized accelerated proximal gradient algorithm, to select small regularization parameters for the cartoon parts, to select big regularization parameters for the texture parts, finally to compose the restoration of the cartoon part and the restoration of the texture part to achieve restoration image. The simulation experiment by MATLAB simulation show that this method is not only converges faster but also better than. the effect of generalized accelerated proximal gradient algorithm, especially suitable for recovering the image of small fuzzy degrees.Content organization is divided into three parts of the paper:theoretical modeling, pre-processing and recovery algorithm. Theoretical modeling is complete image degradation and total-variation modeling; pretreatment complete the blurred image of cartoon texture decomposition; recovery algorithm to solve the image restoration model.In this paper, the arrangement of contents first expounded the background and the practical significance of the image restoration, and second reviewed image restoration studies, and third introduced cartoon texture decomposition of image restoration based on the total variation regularization model, the last to be verified by MATLAB simulation experiment. The specific six chapters arrangements introduction, image degradation and total variation regularization model, accelerated proximal gradient algorithm to restore the image, the image of the cartoon texture decomposition, cartoon texture decomposition combined with generalized accelerated proximal gradient algorithm to restore the image, and Summary and Outlook of study.