A New Method for Video Shot Detection Using Visual Bag Using Corner Detector

In this paper, a system called Video Shot Detection using Visual Bag of Words is proposed for effective shot change detection using multiple feature extraction algorithms with the help of corner detectors. These multiple features which form Bag of Visual Words are extracted in order to retrieve patches from each frame. The detection method was tested on different videos such as sports, entertainment and other documentary videos which contain different types of shots having different objects. Our results have shown that the system can produce accurate shot change detection which is useful for video content analysis and indexing.