A comparative analysis on pixel-based blind cloning techniques

Digital image forensic is an emerging discipline that signifies a never ending struggle against image forgery. In this paper an attempt is made to overview the recent developments in the field of blind forensic detection techniques. The cloning forgery is one in which instead of having an external image as the source, the textured region from the same image is copied and moved to a different location for vanisihing or adding certain feature(s) in an image. We would limit our focus in this paper to the artefacts introduced at pixel level in an image, rather than anomalies introduced at format, camera, geometric and physical based. Two types of pixel based forgery detection techniques exist in literature, the Block based methods and Key-point based methods. Both the methods have their own advantages and limitations. The aim of this paper is to develop an analytical framework which provides comparison on performance & results, to narrow down the techniques which are more efficient, robust and give better evaluation.