The rapid progression of digital technology has led to an increasing number of advantages associated with acquiring voluminous information through internet access. The proliferation of media alteration software has facilitated the manipulation of multimedia data. As a result, the authentication and integrity of images are crucial in numerous disciplines. Image forensics is a burgeoning field utilized to assess the reliability of digital pictures. The purpose of image forgery is to produce obscure images while concealing vital and useful information. Copy-move forgery (CMF) is a form of forgery that exposes both the general public and image forensics specialists to grave danger. For effective forgery detection, the aim of this research is to hybridize discrete cosine transform (DCT) and accelerated KAZE (AKAZE). DCT is initially applied to each block subsequent to block division. In addition, segmentation is performed utilizing k-mean clustering, which is subsequently followed by feature extraction via AKAZE. Following this, the k-nearest neighbor algorithm is utilized to match the characteristics. Ultimately, morphological processes are employed to photographs to expose fabricated regions. The experimental outcomes are implemented on standard datasets for image manipulation. A range of performance parameters, including precision, recall, F1 score, and F2 score, are assessed and demonstrated to be superior in comparison to the currently employed methodologies.