Detection of blue screen based on edge features

Blue screen compositing is an important editing or tampering operation to change the content of videos. In order to identify the authenticity of the video resources, a new approach to detect blue screen compositing is proposed in this paper. By exploring the change of correlation between the color signals on the edge of each object, we can find and locate the composited objects in the video scene. The experimental results show that our proposed algorithm can successfully detect the blue screen compositing in videos with different output bitrates.