A Source Video Identification Algorithm Based on Motion Vectors

With the easy availability of DV (digital video) and associated software to manipulate videos, ascertaining the integrity and authenticity of digital videos has become an urgent and critical issue in today's digital community. In this paper, a new source video system identification algorithm is proposed based on the motion vector information in the encoded stream; it takes full advantage of the various characteristics in the motion estimation algorithm in different video compression systems, and combines a k-nearest neighbor (k-NN) classifier to build a complete video system identification scheme. The experiments show this proposed algorithm can effectively identify video streams which come from a number of video coding systems.