Statistical approaches for passive multimedia forensics

Abstract

With the advancement in the usage of digital multimedia contents, it is necessary to keep all digital contents protected and reliable. Due to availability of high class software editing tools, a novice user can change the meaning of multimedia contents. Such alteration if done for entertainment purpose, it is acceptable. But if alteration in the digital multimedia contents is done to defame any personal or to defame a group, then this is a severe crime. To investigate altered multimedia contents, there is an urgent need of developing some tool which can detect forgeries made. In response to this various work has already been started by the researchers from the multimedia forensic domain. Researcher have followed two approaches to detect multimedia forgery. One is active approach another is passive approach. In active forensic techniques, investigation is done on prior embedded information such as watermarking. If a forger attempts to change a picture/video, he does not bother about the watermark embedded with the contents. In such cases, investigator tries to extract watermark. A distorted watermark confirms forgery. In passive approach, digital multimedia contents does not have any prior embedded information. This approach has higher challenges because investigator has to apply blind approach to detect forgeries. These forensic techniques operate by detecting imperceptible traces left by editing operations in digital multimedia contents. The objective of this study is to actively contribute to this field of research, developing efficient techniques to protect digital contents and verify their integrity.