Robust object removal with an exemplar-based image inpainting approach

Object removal can be accomplished by an image inpainting process which obtains a visually plausible image interpolation of an occluded or damaged region. There are two key components in an exemplar-based image inpainting approach: computing filling priority of patches in the missing region and searching for the best matching patch. In this paper, we present a robust exemplar-based method. In the improved model, a regularized factor is introduced to adjust the patch priority function. A modified sum of squared differences (SSD) and normalized cross correlation (NCC) are combined to search for the best matching patch. We evaluate the proposed method by applying it to real-life photos and testing the removal of large objects. The results demonstrate the effectiveness of the approach.