Video Forgery Detection Using Natural Scene Statistics

Abstract

Images and videos have become an integral part of our every day life. The availability of image/video processing software has made it even more easier to change the entire meaning and context of media fles. These digital media fles can easily be tampered with, manipulated and synthesized using various softwares by an individual or an organization for their personal gains. Many approaches for the detection of video forgery are evidently present in literature, but all of them have the same principle objective to detect and locate the forgery. One of the most primitive, yet critical type of detection technique is the Copy-Move forgery detection (CMFD). In this thesis, an in depth study has been conducted of the techniques that are already present in literature whilst highlighting the limitations and challenged faced. We have proposed a completely Blind (Passive) video forgery technique that extracts natural scene statistical features in spatial domain.Both GGD and AGGD based features have been computed and are ftted using the MVG model. The features include mean, standard deviation, local sharpness and gamma in four orientations along with βl andβr The proposed methodology for forgery detection yields better results for publicly available SULFA and VTD databases. Hyper-parametric tuning of the Random Forest Classifer has been used for classifcation purpose. The F-Measure score (FM), recall, precision and accuracy are calculated for the proposed algorithms which are then compared with the algorithms already present in literature in order to make a fair evaluation. The FM score achieved is 0.98 while the accuracy is 99% after k-fold cross validation.Further enhancement of the deep learning and learning transfer approach is recommended for future work.Inter Frame detection and pixel-wise detection shall be done in future.