An Image Forensics Analysis of Decision Fusion Approaches

A Forensic Image is a is often accompanied by a calculated Hash signature to validate that the image is an exact duplicate of the original which is mainly focus on detection of artifacts introduced by single processing tool. Hence making it necessary for developing several for detection of artifacts. It is by introducing theoretical frameworks, based on Dempster-Shafer’s Theory of Evidence, Fuzzy Theory and on Bayesian inference respectively. These decision fusion theories are mainly of heterogeneous or having the conflicting outputs of forensic algorithms. These models are easily expandable to an arbitrary number of tools do not require output to be probabilistic and take into account available information about tools reliability. To validate the proposed approaches, some experiments addressing a simple yet realistic scenario in which three forensic tools exploit different artifacts introduced by double JPEG compression to detect cut and paste tampering within a specified region of an image. The results we obtained are encouraging when we compared with the performance of a simple decision method based on the binary OR operator.