Parallel Image Forgery Detection Using FREAK Descriptor

Nowadays, a large amount of information is being exchanged in the form of images. This information can be tampered easily through a process called forging. This paper focuses on detection of copy-move image forgery in an image. To implement it in a faster way, parallel copy-move image forgery is proposed. The features from accelerated segment test (FAST) method is applied to detect the key points of the input image. After detection of keypoints, fast retina keypoint (FREAK) binary descriptor method is used to find the features of these keypoints. These features are then matched, and the correlation factor is found to detect image forgery. The image is split into various regions, and detection in each region is done in parallel. Hence, it helps to find out the image forgery in a faster way. The analysis shows that the proposed method is performed in a faster manner to detect the forged region.