Inpainting Of Ancient Mural Cracks Based On Compression Sensing And Image Decomposition Technology

Mural is an important material data for researching on ancient Chinese religious belief,customs,painting and so on,and it is a valuable non-renewable resource.However,since the impact of natural weathering and man-made destruction,the murals are damaged by various diseases.With the fast development of the computer image processing technology,digital image inpainting technology has been applied increasingly to the cultural relic protection,the digital inpainting of murals greatly reduces the irreversible risk that caused by artificial protection.Since the large number of mural high-definition image storage,the traditional inpainting algorithm is less efficient.The compressed sensing(CS)theory which fully considers the sparsity of the signal is used to inpaint,and can improve the efficiency of the murals inpainting.In this paper,take the tang dynasty tomb murals "polo figure" as the research object.Combined with the thought of the overall structure and the local texture of ancient murals in artificial restoration and the characteristics of ancient murals,meanwhile since the existence of ancient murals in a large number of many cracks,an ancient murals cracks inpaiting algorithm based on compressed sensing and image decomposition is proposed,and an intelligent terminal system is designed for the cracks inpainting.Specific work is as follows:(1)By analyzing the characteristics of the tang dynasty tomb murals,the morphological component analysis(MCA)method based on sparse representation is used to inpaint the image.For the traditional MCA algorithm is difficult to define the number of iterations,the iteration termination condition is modified and the convergence speed is accelerated.Then,according to the complexity and sparse of cartoon and texture,the improved total variation(TV)algorithm and the K-Singularvalue decomposition(K-SVD)algorithm are used to inpaint the image.The experimental results show that the proposed algorithm which take into account the effect of the texture and structure,and can improve mural cracks inpainting efficiency and accuracy.(2)Combining with the inpainting algorithm proposed in this paper and the characteristics of the ancient murals,an ancient mural intelligent terminal repair system was built.The system includes multiple modules,such as login registration,gallery browsing,crack marking,image decomposition,image inpainting and image evaluation,which can mapping the crack disease,decomposing image and inpainting,and it brings convenience for the researchers who can use a mobile device to inpaint ancient murals.