Enhancement of Media Splicing Detection: A General Framework

Digital media (i.e., image, audio) has played an influential role in today information system. The increasing of popularity in digital media has brought forth many technological advancements. The advancements, however, also gives birth to a number of forgeries and attacks against this type of information. With the availability of easy-to-use media manipulating tools available online, the authenticity of today digital media cannot be guaranteed. In this paper, a new general framework for enhancing today media splicing detection has been proposed. By combining results from two traditional approaches, the enhanced detection results show improvement in term of clarity in which anomalies are more explicitly shown, providing easier and faster way for a forensic practitioner to investigate and verify the authenticity of the target digital media. Regarding the experiment, the developed framework was tested against a number of realistic tampered (spliced) media. Moreover, the enhanced detection results are compared with traditional approaches to ensure the efficiency of our proposed method in the realistic situation