An Image Inpainting Technique Based on 8-Neighborhood Fast Sweeping Method

Image inpainting is an important element in image restoration study, whose purpose is to restore the lost information automatically based on existing information. Image inpainting can be used to restore damaged images and videos, remove text and hidden errors on videos, and so on. Although the basic idea of the image inpainting is very simple, many of the image inpainting algorithms are very complex and difficult to achieve. In order to get a fast algorithm, this paper proposes a new algorithm based on 8-neighborhood fast sweeping method and improves the algorithm in details. By inpainting the same image using Bertalmio´s method and our method, the experimental results verify that our algorithm is fast and feasibility.