Histogram-based perceptual hashing for minimally changing video sequences

An increasing number of multimedia applications require techniques for efficient and automatic content identification and verification. Here, perceptual hashing technologies apply. Due to the diverseness of video data, algorithms whose design considers the properties of the different video classes are required. Therefore, different approaches exist to extract a perceptual hash. In this article, we propose a histogram-based perceptual hashing algorithm for minimally changing video sequences. The algorithm is described and the performed evaluation shows improvements of the identification performance for minimally changing video content.