Inpainting Images on Implicit Surfaces

Planar image processing has been widely investigated for many years. The processing operations include denoising, edge enhancement, edge detecting, inpainting, and others. But there exists little work about processing images on surfaces, since it is difficult to extend the classic methods to deal with the problem. In this paper, we study the inpainting algorithm of images on implicit surfaces based on the method of energy minimizing and PDE. It’s a generalization of the inpainting algorithm of planar images. An intrinsic energy functional is defined over surfaces. Energy minimization problem is solved by a numerical method, which needs data extrapolating. Another contribution of this paper is a theorem on how to control data extrapolating for processing images on implicit surfaces. The experiment results show the efficiency of our method.