An Ontology of Digital Video Forensics: Classification, Research Gaps & Datasets

The world lives in an epoch where people have an incredible variety of digital multimedia technologies. Although historically it was established that the computerized information is authentic information, but the technology of today has begun to erode this assuredness because an opponent uses several distortions to edit the original video for trademark infringement. Many investigators have introduced video forgery detection systems, due to the advent of omnipresent digital videos on the internet and to overcome the difficulties. This paper depicts a comprehensive overview of different video forgery detection techniques. It specifically discusses the importance of the techniques by classifying them according to functionality. This study delivers an authoritative and thorough review of existing digital video forgery detection techniques and also highlights various research gaps of the current techniques. Furthermore, the existing video forgery detection techniques are assessed by analyzing accuracy, precision and recall. Finally, the paper discusses different public and private datasets available for evaluating video forgery detection techniques.