Content-based video copy detection using spatio-temporal compact feature

This paper presents a content-based video copy detection algorithm that detects online distribution of illegally copied video. Particularly, the proposed algorithm uses keyframes with abrupt changes of luminance, then extracts spatio-temporal compact feature from keyframe. Comparing with the preregistered features stored in the database of videos, the proposed approach distinguishes whether an uploaded video is illegally copied or not. Note that we analyze only a set of keyframes instead of an entire video frame. Thus, it is highly efficient to detect illegal copied video when we handle a vast size of videos. Also, we confirm that the proposed method is robust to a variety of video modification that are often applied by online video redistribution, such as aspect-ratio change, logo insertion, caption insertion, visual quality degradation, and resolution change.