Research And Application Of Exemplar Block-Based Image Inpainting Algorithm

Image inpainting is a new pop area of image restoration. It is recently regarded by many image processing researchers at home and abroad. Image inpainting is to fill in the image information on a blank domain based on the information available outside, essentially, it is an interpolation problem. It has been widly applied in restoring scratched oemld photos and old movies, text removel from images, digital zoomings and especial effects in movie, now it is also applied in restoring reality photoes from which stumbling block is roved.This paper introduces the particularity of image inpainting models, discusses the PDE-based image inpainting models, the texture- synthesize-based image inpainting models and the PDE-texture-synthesize-based image inpainting models. The algorithms corresponded to the models are also discussed. An exemplar block-based image inpainting algorithm is achieved after the comparison of the applicability, merit, and disadvantage of the models.First, in the exemplar block-based image inpainting algorithm, the priorities of the inpainting blocks on the"fill front"are calculated in order to determine the block with the highest priority. Then the best exemplar block corresponding to it is discovered in the exemplar region, and the relevant data is copied from the best examplar block to the unfilled parts of the block with the highest priority. Finally, the confidence values of pixels which have been just filled are updated. The whole damaged region will be fully filled through iterating the above steps.In the exemplar block-based image inpainting algorithm, the confidence of the synthesized pixel values is propagated in a manner similar to the propagation of information in inpainting. The actual pixel values are computed using exemplar-based synthesis. The simultaneous propagation of the texture and structure information is achieved by a single efficient algorithm. Computational efficiency is raised by a block-based sampling process.Experimental results indicate that the method can effectively to fill the texture and structure of the damaged image areas simultaneously, combine the block matching and the filled order well, effectively play the filling propty role in the inpainting process, and reconstruct the damaged image more perfect. The exemplar block-based image inpainting algorithm retains not only the merit of pixel gray level optimization but also the continuity of structure diffusion, achieves a balance between them by combining the data value and the confidence. The algorithm can also be used to remove undesired objects on digital images and filled the blank areas. The challenge is to fill in the hole that is left behind in a visually plausible way.