A Fast, Block Based, Copy-Move Forgery Detection Approach Using Image Gradient and Modified K-Means

摘要

In recent years, due to the fast development of digital images, a rapid growth of research interest in the forgery detection in digital images has been happened. One of the most common techniques in creating forged images is copy-move (region duplication) technique. In this paper, a new method for copy-move forgery detection in digital images is proposed. In this paper a region duplication detection technique which utilizes the image gradient is proposed. In the proposed approach, first the gradient of image is divided into overlapped blocks. Using gradient versus other techniques, decreases processing time in feature extraction step. A fast pre clustering algorithm is another added step to speedup method by dividing search area into some subset. The unknown parameters of proposed method are determined by implementing different conditions on two standard databases. Finally, the performance of the proposed method is compared with some state of art methods and the acceptable accuracy and lower run time of it, is verified.