Image in-painting techniques - A survey and analysis

Abstract

Digital in-painting is relatively a young research area, yet a large variety of techniques were proposed by the researchers to correct the occlusion. Image in-painting aims to restore images with partly information loss and tries to make in-painting results as these missing parts in such a way that the reconstructed image looks natural. Many different types of image in-painting algorithms exist in the literature. However no recent study has been undertaken for a comparative evaluation of these algorithms to provide a comprehensive visualization. This paper compares different types of image in-painting algorithms. The algorithms are analyzed in both theoretical and experimental ways, which have made the suitability of these image in-painting algorithms over different kinds of applications in diversified areas.