Image Manipulation Detection using Intrinsic Statistical Fingerprints

The digital images or videos are becoming main part in the field of information forensics and security and there uses are increased, so it is must to create forensic techniques capable of detecting image or video frame alteration operations and fo rgery image. In forgery images and videos, there are number of image processing operations, such as contrast enhancement, histogram equalization, speckle noise, image scaling. We present the forensic methods for detecting globally and locally applied contra st enhancement and method for detecting histogram equalization in image by identifying the features of each operation's intrinsic fingerprint. Additionally we propose a method for detecting noise as well as the method for detecting image scaling or croppi ng by observing the intrinsic fingerprint of specific mappings. Finally, we test the efficacy of each proposed forensic technique.