A Novel Forgery Detection Algorithm for Video Foreground Removal

Video processing software is often used to remove specific moving foreground from a video. Existing forgery algorithms for detecting this type of tampering generally suffer from inefficiency and are not effective for the forged videos under complex background. To address these problems, we propose a novel forgery detection algorithm for detecting video foreground removal. The algorithm first calculates the energy factor (EF) of each frame to identify forged frames. An adaptive parameter-based visual background extractor (AVIBE) algorithm is then designed to detect suspected regions from the forged frames determined in the first stage. After eliminating false detection by calculating the difference of EF between suspected regions in the forged frames and the corresponding regions in the authentic frames, the algorithm finally locates the tampering traces. The experimental results show that our proposed algorithm has higher computational efficiency and accuracy as well as better robustness than those of previous algorithms.