Author : Mariam Saleem

Digital image have its significance in several applications from financial documents, scientific literatures /journals, medical records, journalism or evidences in a court of law. However, with the mushroom growth of handy, inexpensive editing software the truthfulness or uprightness of image can no longer be taken for granted. So there is a need to establish techniques in order to declare the integrity of digital images. Digital Image Forensic has somehow solved this problem. The main objective of this field is to develop detection technique to estimate the authenticity of images and to reveal the possibility of image being forged. A wide range of forgery detection techniques have been established in the recent years. Blind Forensic Detection technique-cloning is one of the most common forgery methods. In this paper an attempt is made to develop a robust and effective key-point based forgery detection approach. In this work first SIFT is employed to find image key-points (Location which carries definite details of image) and to extract image features at the detected key-points. For effective matching and detection a productive algorithm called Multi Hop Jumping is then utilized to jump over unnecessary or false key-point matching. A rational framework is developed at the end by providing relative results using previous key-point based forgery detection techniques to witness the proposed technique which is not only efficient, reliable, effective but can also reduce processing time as compared to other key-point based techniques.

Keywords: Digital Forensic, Copy-Move Forgery, Cloning, Pixel-Based, Key-points, SIFT

Article published in International Journal of Current  Engineering  and Technology, Vol.4,No.4 (Aug- 2014)