Variational Model For Image Restoration Split-bregman Algorithm

Image inpainting is an important part of image science. Digital image inpainting is researched focus on the field of machine vision and computer graphics. Also it has a wide range of applications and has a significant place in the image processing. The objective of image inpainting is to fill in its lost part according to the surrounding information. And restoration of the current research is related to images of both repair and synthesis. In this paper, around the partial differential equations to image processing, the principle, methods and applications of image inpainting have been studied and discussed. Because of the low calculate efficiency of the image inpainting algorithm, recent years, a new algorithm named Split Bregman algorithm has been extensive research and application. This article focuses on the classical partial differential equations in image inpainting and discusses their fast algorithm, Split Bregman algorithm. The main work and innovations include the following:Firstly, the existing mathematical model of image inpainting and their algorithm were in-depth studied, and the advantages and disadvantages of these models were deep analysis and discussion. Secondly, the Split Bregman algorithm to be used in the classical image inpainting models, compare the speed of operation and the results compared. Thirdly, this algorithm was used in some color coupling image inpainting models, and compares the results of their compute results.In summary, this paper in-depth research was made in the field of digital image inpainting, and made some valuable research results. Digital image restoration is still a more promising area of research, it is valuable for further studies.