Image tampering detection in image forensics using earthworm-rider optimization

This article introduces a novel earthworm-rider optimization algorithm (EW-ROA), by integrating the earthworm algorithm in the rider optimization algorithm for the detection of the tampered regions optimally. Initially, the individual binary maps are generated using the image blocks of the input image are fed to five forensic tools, and then they are concatenated into a single binary map. The features are extracted from the binary map moreover they are feed to classifier for the detection of the tampered image via the deep belief neural network, which is trained using the proposed earthworm-rider optimization algorithm. The proposed system attains a highest accuracy of 0.9402, sensitivity of 0.98, and specificity of 0.98 that shows the superiority of the proposed system in effective tampering detection.