Abstract
Internet use has become a staple of our lives to transfer data. Although anyone can alter or modify these data, as image modifying tools are readily available. The slightest change in the medical image can lead to a wrong diagnosis, and the vital goal is not achieved. Tamper detection has been proposed as a solution for ensuring the integrity and authenticity of digital images. The goal is to identify and pinpoint any image changes that have taken place after transmission. An innovative watermarking technique is presented in this paper for detecting tampering. The approach relies on the Code Division Multiple Access (CDMA) technique. In order to generate the Common Code (watermark) data, the blocks must be encoded, followed by a Walsh table. This embedding process occurs in the LSB of each block. During the semi-blind detection step, the encode (data representation) matrix is necessary during the verification and detection steps. The proposed approach evaluated various types of attacks and tampered processes. According to the results of the experiments, the proposed approach outperforms existing and related approaches in terms of quality, imperceptibility, capacity, and tamper detection. The proposed method is evaluated using the Stirmark benchmark and CASIA v2 dataset. The approach provides a high quality and there is no degradation after embedding the watermark in the original image. Also, the watermarked approach is imperceptible where the PSNR is 56.4 dB and the SSIM is 0.9983, it consumes less space in the original image, its bPP is 0.85. In addition, it requires less computational resources than other approaches with an embedding and extraction time of 3.798 s. Further, the detection performance of the tampered data is 99.965. In addition, the approach provides promising results against spliced tampering, with an average F1 score of 0.8969.