Copy-move forgery detection using a bat algorithm with mutation

One of the challenges today is capturing fake images. One type of image forgery is a copy-move forgery. In this method, a part of the image is copy and placed at the most similar point. Due to the existing algorithms and processing software, it is not easy to detect forgery areas and has created challenges in various applications. Based on the Bat algorithm, the proposed method has tried to help detect fake images by finding forgery areas. The proposed method includes a simple image segmentation and detection of forgery areas with the help of the BAT Algorithm with Mutation. According to the proposed algorithm, the image is first grayed out, then divided into 100 pieces. The optimal Bat algorithm randomly selects some component image and performs a similarity search. The mutation operator is used to avoid getting stuck in the local optimization. The proposed algorithm does not get stuck in the local optimization with the help of the mutation operator and can find forgery areas with a precision of about 81.39% for the IMD dataset and about 81.04% for the MICC-F600 dataset.