Detection of Malicious Video Modifications using Perceptual Video Hashing

The perceptual video hashing (PVH) function produces fixed-length and a compact video fingerprint (or perceptual video hash) based on the perceptual contents of the video. This fingerprint can be used for detection of malicious video modifications or video tamper detection based on its perceptual contents, video authentication and video copyright protection. Here, the video fingerprint is extracted using the three-dimensional (3D) radial projection (RP) technique and the two-dimensional (2D) discrete cosine transform (DCT). The proposed algorithm is inspired from the following algorithms: video hashing based on 2D-RP of key frames and the video hashing based on 3D-RP technique. The performance of all the aforementioned algorithms are assessed using the receiver operating characteristics curves. The proposed algorithm perform better in the context of detecting malicious attacks.