Enhanced image forensics using dynamic authentication and recovery with CRC and chaos mapping

Abstract Due to the huge advancement of various image editing software, it is very challenging to ensure the authenticity and integrity of a digital image. In this paper, an innovative technique is proposed for image forensics, i.e., the detection of tampered images and their recovery using watermarking and chaotic mapping techniques. The proposed algorithm effectively locates the tampered areas and achieves better restoration of the original image through hidden information. The chaotic system provides the mapping between the sub-blocks of an image to maintain randomness of restoration and authentication information and forms a pseudo-random cyclic chain. If the image gets tampered with, then the chaotic system can again rebuild the original image using the recovery information hidden in the eighth neighboring blocks. The method ensures strong preservation of image integrity, even under extensive tampering, while maintaining high recovery fidelity. Experimental results demonstrate that the proposed approach exhibits superior performance in detecting complex tampering scenarios, such as collage attacks, and significantly improves the visual quality of the recovered image.