A Review of PDE Based Local Inpainting Methods

Abstract

Image inpainting is the process of recovering the damage areas in the images in an undetectable way, it is considered the important one of the subjects in image processing. There are many applications of image inpainting include the restoration of damaged images, paintings, and movies, to the removal of selected objects, such as text, lines, subtitles, publicity, and stamps. The main objective of inpainting is to reconstruct the missing region in such a way that the observer does not come to know that the image has been manipulated. Inpainting methods can be categorized into global and local methods, the global methods are applied to reconstruct the damaged areas in the image based on the information in the data of images that have the same content. While the local methods are used to reconstruct the missing regions based on the information in the rest parts of the image. There are several local methods proposed for image inpainting such as PDE-based inpainting (PDE-BI), exemplar-based inpainting (EBI), hybrid, and texture synthesis methods. In this paper, a review of different PDE and variational methods used for image inpainting is provided. Different PDE-BI methods like 2nd-and high-order of variational and PDE methods are discussed with its pros and cons.