An efficient recording method for video recorders using the video signature tools

The Video Signature is a high-performance content fingerprint that detects the same video streams with high accuracy. It can be applied to video recorders in order to reduce data size can by deletion of the same parts. However, it requires a large amount of computations. In the proposed method, using sound features, candidates of commercial messages are extracted. The video identification is done only for candidate segments. The experimental results show 63 % reduction in processing time.