摘要
提出了一种新的脆弱盲水印生成算法用于图像完整性证明和内容篡改证明。首先将原始图像分块,对各分块进行一次整数小波变换;计算4个子带(LL,HL,LH,HH)小波系数绝对值的均值,分别映射为混沌初值,经过迭代、量化生成二值水印信息;最后将水印信息随机嵌入到4个子带中。实验结果表明,所提出的算法提高了篡改检测率,对图像的篡改具有高度敏感性和准确的篡改定位能力。水印提取为盲提取,篡改检测只用到了篡改后的水印图像。
A new fragile blind watermarking algorithm for integrity verification and tampering proof of image is proposed. Firstly, we divided the original image into some blocks, operated integer wavelet transform for each block once, and worked out the average of absolute value of wavelet coefficients in four sub-bands(LL,HL, LH,HH)respectively. Then we changed each average into the initial value of chaos. The watermark information was obtained by iterating of chaos and quantifying the chaos sequence. At last, we embedded the watermark information into four sub-bands randomly. Ex- perimental results show that our algorithm improves the probability of tampering test further and has high sensitivity to image tampering. It can exactly locate the tampering part also. In addition, the watermark information can be extracted without original image, watermark image, and tampering test can be done only with the watermark image tampered.
出处
《微电子学与计算机》
CSCD
北大核心
2007年第11期127-130,共4页
Microelectronics & Computer
关键词
混沌映射
整数小波变换
篡改证明
脆弱水印
chaotic map
integer wavelet transform
tampering proof
fragile watermark