Video Inpainting Using Seam Carving Method

Abstract

Video became something important of our daily lives and in many areas such as surveillance cameras and mobile cameras therefore it has attracted considerable attention in multimedia and computer vision researches in recent years. One such research is video inpainting.Videoinpainting aims to restore missing or tainted regions or remove objects present in a video sequence from neighboring scenes. The proposed method in video inpainting is object removal technique based on seam carving algorithm, where the static AVI video type and object dynamic are used and working video for different number of frames per-second with resolution (320 pixel by 240 pixel) and (640 pixel by 480 pixel) .The implementation in the proposed seam carving algorithm for object removal is obvious that discrete Laplacian edge detection filter gives the highest matching with four videos, which proves that its work is the best in detecting the energy of a video, and the best accuracy for whole system is (86.57516).