摘要
提出一种基于元胞自动机变换的可恢复图像认证算法,将图像分割成不重叠子块,对每个子块进行两层二维元胞自动机变换CAT,从第二层低频系数生成子块认证水印和恢复水印,分别嵌入当前子块和对应子块的第一层低频系数,形成含水印图像;用认证水印进行图像认证;用恢复水印恢复被篡改图像子块。实验表明,该算法安全性高,且具有很强的抗VQ攻击能力。
A restorable image authentication with cellular automata transform is proposed. The original image is splited into non-overlppin blocks and two layer cellular automata transform （CAT） is performed on each block. Authentication watermark and restoration watermark are produced from the second-layer low frequency coefficients and embedded into the first-layer low frequency coefficients of the current block and the corresponding block separately. Image authentication is played with the authentication watermark, and the tampered blocks can be restored with restoration watermark. Experiment results show that the scheme＇s security is strong and can resist vector quantization attack effectively.
出处
《电子科技大学学报》
EI
CAS
CSCD
北大核心
2009年第6期1026-1030,1056,共6页
Journal of University of Electronic Science and Technology of China
基金
四川省青年科技基金(06ZQ026-054)
关键词
元胞自动机变换
图像认证
图像恢复
安全
篡改定位
cellular automata transform
image authentication
image restoration
security
tamper location
