Application Of Image Inpainting Algorithm In Chinese Painting

Chinese painting is one of our traditional treasures of art, which has made a great contribution to the heritage and development of the Chinese civilization in the long history. However, a lot of paintings have been more or less damaged because of the long time or improper storage, transportation, damp and other factors. In this text, we analyze and sort the problems of damage, and find out some principles and method of image inpainting.For the problem of crease inpainting, we use an improved inpainting algorithm based on the exemplar filling. After getting the target region being designated by user, the algorithm will automatically finish the inpainting work. To achieve much better effect, we use image segmentation to guide the searching process of the best exemplar, and find a method called GraphCut to get a seamless paste. After that, we use artifact detection to detect the target region, in order to optimize the algorithm.For the problem of polluted background inpainting, we use image interpolation, that user select a guided field , which is not polluted, to fill the target region that user want to. We try to get a seamless boundary of target region by Poisson equation, after the image interpolation, and to find the result of Poisson equation we use a combined algorithm of Steepest Descent and Conjugate Gradient. We also judge the area of the guided field, if it is smaller than the area of target region, we expand the guided field before interpolation.The result of experiment improved that the algorithms we used can repair the problems of crease and background pollution in paintings, and get satisfying effect in visual.