Image Forgery Detection using Statistical Moments
Information security is inevitable and indispensable in today’s world. It is found that we have more interactions in online. The threats involved in these transactions make information security necessary. Information security intends to protect the content getting revealed and/or modification of content. Attacks are categorized as active and passive attacks. For active attacks changes the content. Hence security measures against active attacks involve detection of attack. Modification of content generally involves copy-move-forge attack, image re-sampling and image forgery through splicing. This paper deals with identification of image forgery using statistical parameters of the image. Hu moment features are extracted from the image and are used for detection of forgery. In addition local features and global features are used for detection of forgery in the image. Insertion attack, deletion attack and copy-move-forge attack are detected and the detection performances are compared with Difference expansion based methods and evaluated.
Keywords- copy-move-forge attack, statistical moments, feature vector