Abrupt cut detection based on mutual information and motion prediction

In this paper we propose a novel method of abrupt cut detection. Abrupt cut is the most frequent kind of transition between shots. Our method improves the existing method based on the mutual information. At first we add the motion prediction to each frame and then calculate the mutual information. The obtained results of our method were compared with the ones from the original method without the motion prediction. Our approach provides higher robustness to object and camera motion.