Adaptive Tamper Detection and Self-Recovery of Medical Images Based on Image Inpainting
Mingyue Li, Jianzhao Li, Guangwei Liu, Ziqian Liu, Xiangdong Tian, Shaomin Zhang · 2024
This paper introduces a novel self-recovery mechanism for tampered medical images based on image inpainting. During the watermark embedding phase, the algorithm proposes to embed the lesser watermarks in the contour edge region of the image. In the stage of recovering the tampered region, the algorithm proposes to recover the tampered region with the help of an image inpainting algorithm that transforms the tampered block matching problem into a minimization problem of non-local self-similar image blocks. Experimental results demonstrate that the average PSNR of the watermarked image is more than 50 dB for a watermark capacity of 0.015625. Additionally, when the tampering rate is below 20%, the average PSNR of the restored tampered medical image surpasses 40 dB.