Two-order Wavelet Transform-Based Dual-Tamper Detection and Self-Restoration

In this research, we propose a neat dual-tampering detection approach to further retrieve the forged content of 1080p images. The recovering of actual information is derived from the low-frequency components of the second-order wavelet. The procedure initiates with the segmentation of the image into a non-underlying section. To make better the quality of the recuperate image, a radially symmetrical mapping technique is employed to safeguard the recovery bits. The most notable bit-plane of each segment is utilized to generate the first component authentication key through the MD5 algorithm, while the second component authentication key is produced using SHA1 based on the recovered data. Subsequently, the watermarked image is formed by embedding the recovery bits and the dual authentication keys into the least significant bits of the original image. During the tampering detection phase, the calculated authentication keys are compared against the retrieved keys. Any discrepancies result in the block being classified as tampered. Following this, centrosymmetric mapping is applied to extract the low-frequency coefficients of the second-order wavelet for the purpose of image self-recovery. The inverse wavelet transform is then executed to restore the tampered regions with the recovery data, thereby finalizing the self-recovery process. Our experimental findings indicate that this approach yields enhanced recovery quality and more effectively conceals the embedded information.