IMAGE AND VIDEO INPAINTING

Abstract

Inpainting, the technique of modifying an image and video in an undetectable form, is ancient as art itself, has received considerable attention in recent years. The goals and applications of inpainting are numerous, from the restoration of damaged paintings and photographs to the removal/replacement of selected objects in images. Video inpainting technique has many applications including surveillance video processing, and video editing during post-production in the movie industry, etc. In the case of video, one important task in inpainting is to repair a moving object/human. In this dissertation, we tried to examine the most appropriate image and video inpainting techniques. Although the inpainting basics are straightforward, most inpainting techniques published in the literature are complex to understand and implement. We used a novel algorithm for digital inpainting of still images that attempts to replicate the basic techniques used by professional restorators. After the user selects the regions to be restored, the algorithm automatically fills-in these regions with information surrounding them. We have also used a novel post processing technique for completion of static background and moving foreground that comprises extending of image inapinting to video inpainting while maintaining temporal consistency. For the knowledge of background and foreground objects we also used a background foreground segmentation technique. The results obtained in our experiments were quite promising.