Keypoint Based Tampered Image Identification

Identifying the tampered regions within an image is a challenging field of study. Images are often altered using editing programs or tools such as Picasa, Photoshop, and others, intending to mislead people or conceal information. Therefore, it is essential to verify the validity of photographs to determine whether they have been altered before extracting any valuable information from them. One specific type of tampering, where an object is intentionally duplicated in the image, is known as copy-move forgery. Based on locating AKAZE keypoints in an image, copy-move forgery area detection has been presented in this paper. The AKAZE features are extracted from the image, and brute force matcher technique is used to match features. In order to group geographically confined keypoints that identify cloned regions or forged portions in a picture, the Birch clustering algorithm is applied to the keypoint locations. Images from the MICC dataset series and CoMoFoD datasets are used for the tests, and the outcomes are compared with those of previously published methods.