Image forgery detection using feature based clustering in JPEG images

JPEG images are most commonly used in a wide variety of applications. JPEG compression properties can be used for forgery detection in digital images. While performing an intended forgery, the image has to be recompressed. Therefore identifying the traces of recompression can be good clue for detecting manipulation. In this paper, a method to detect forgery in JPEG image is presented and an algorithm is devised to classify the image blocks as forged or non-forged based on this classification. The method produces better results than the previous methods which use the probability based approach for detecting forgery.