The Design and Application of PHABS: A Novel Benchmark Platform for Perceptual Hashing Algorithms

Numerous perceptual hashing algorithms have been developed for identification and verification of multimedia objects in recent years. Various application schemes have been adopted to provide the ability of forensic tracking against the illegal use or distribution of multimedia content. Developers and choice-maker are looking for a benchmark tool to compare and evaluate their ongoing algorithms or technologies. In this while, a novel benchmark platform is desired. PHABS provides an open framework to let its users define their own test strategy, perform tests, collect and analyze test data. With PHABS, various performances of algorithms can be tested, and different algorithms or algorithms with different parameters can be evaluated and compared easily.