新颖的内容自适应图像篡改检测与自恢复算法研究  

Novel content-adaptive tampering detection and self-recovery algorithm study

作　　者：李明月 毕秀丽 刘波 LI Mingyue;BI Xiuli;LIU Bo(School of Computer Science and Technology,Chongqing University of Posts and Telecommunications,Chongqing 400065,P.R.China)

机构地区：[1]重庆邮电大学计算机科学与技术学院,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2024年第4期797-806,共10页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金项目(62172067,61976031);国家重点研发计划(2019YFE0110800,2016YFC1000307-3);国家重大科研仪器研制项目(62027827);重庆市教委科学技术研究项目(KJQN202200635)。

摘　　要：针对在原始图像中嵌入大量的水印(认证水印和参考水印)容易造成图像失真的问题,提出嵌入少量参考水印的方法。为了减小图像失真,提出将少量的参考水印信息嵌入在原始图像的边缘、轮廓等区域的方法。在恢复篡改区域时,提出利用图像修复算法将篡改块匹配问题转化为非局部自相似图像块的最小化问题,复制攻击图像本身的相似信息来恢复篡改区域。实验结果表明,在水印容量为0.015626的情况下,该算法得到嵌入水印后图像的峰值信噪比(peak signal to noise ratio,PSNR)均值超过60 dB,恢复图像的PSNR均值超过40 dB。In order to address the problem that embedding a large number of watermarks(authentication watermark and reference watermark)in the original image can easily cause image distortion,we propose to embed a small amount of reference watermark.To reduce the image distortion,it is proposed to embed a small amount of reference watermark information in the edges,contours,and other regions of the original image.In recovering the tampered region,it is proposed to transform the tampered block matching problem into a minimization problem of non-local self-similar image blocks using an image restoration method,and then replicate the similar information of the attack image itself to recover the tampered region.The experimental results show that the algorithm obtains the mean PSNR of the embedded watermarked image over 60 dB and the mean PSNR of the recovered image over 40 dB at a watermark capacity of 0.015626.

关 键 词：水印 图像恢复 图像修复算法 篡改检测 自恢复 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]