基于图像内容和混沌系统的图像认证算法

Image Authentication Algorithm Based on Image’s Content and Chaotic System

提出了一种利用离散余弦变换(Discrete Cosine Transform,DCT)系数和混沌系统的可用于检测图像篡改定位的脆弱性数字水印算法.首先将图像进行不重叠的分块,分块后的图像经过DCT变化后,利用中频系数产生水印.同时利用混沌系统构成图像块间信息嵌入和提取的对应关系,并将水印嵌入到相关块的最低有效位.图像篡改检测过程为:将提取的水印和生成的水印进行比较,通过差异的位置对其进行定位.实验结果显示,该算法具有良好的定位精度,混沌系统的采用,使算法具有较高的安全性.

A novel image authentication algorithm based on DCT transformation and chaotic system is proposed.The host images are divided into non-overlap blocks,then DCT is performed on each blocks.The middle frequency coefficients are used to generate the block watermarks.The relationship between blocks is decided by chaotic system,and the watermarks are embedded into the LSB of corresponding block.The taper region is decided by the difference between the original watermarks and the extracted watermarks.The experimental results show that this algorithm has good tamper localization and high security.