2PASCD: An Efficient 2-Pass Abrupt Scene Change Detection Algorithm

Abstract

Scene change detection in videos is a primary requirement of video processing applications used for the purpose of generating data needed by video data management systems and digital rights management (DRM) systems. Scene change data can be used in DRM systems for effective IPR protection by means of watermarking and fingerprinting selected scenes. In this paper, we present a 2-pass algorithm for abrupt scene change detection in videos based on x2 color histogram comparison of subwindows of successive frames and sliding window detector method. In the proposed method, we improve the efficiency of scene change detection by a second level inspection of the detected scene changes using a fixed average and a moving average to ascertain whether the detected scene boundary is correct or the scenes need to be combined. This method gives precision and recall values better than any existing method.