Two Fold Image Forgery Detection System Using Combined Key point based method and Block based method

Abstract

In this paper to detect image forgery, two standard methods, Key point based method and block based method, are combined to form a two fold method which gives more accuracy. This method first use segmentation method that is used in block based method. Features are extracted on these small patches of segmentation; by means of key points based method. Then these patches are compared with each other and distance is measured between them. If two patches have same distance then a region of image is copied and paste on the same image and the image is forged. This process is done in two segments to ensure good and accurate detection. A good and reliable working of the algorithm is ensured by comparing the proposed algorithm with some state of the art standard public available datasets.