摘要
利用DWT系数的周期特性,结合移位置乱和CDMA技术,提出了一种在分块DWT系数中隐藏灰度水印图像的算法.该算法首先将载体进行16×16分块,对每小块进行多级小波分解,然后对水印分组,嵌入载体,最后进行IDWT变换,得到嵌入水印的图像.实验结果表明,该算法在无攻击时可以实现完全盲提取(NC=1.0000),对图像剪切、JPEG有损压缩、中值滤波等攻击,有较强的鲁棒性(NC>0.5).
Through using the periodic property of DWT coefficients, the digital image scrambling transformation namely Shift and CDMA; this paper proposes a grayscale image watermark embedding algorithm on block discrete wavelet transform DWT coefficients. Firstly, the host image is partitioned into block size of 16 × 16, every block is decomposed of multilevel wavelet, then watermark is packeted, and hidden, finally, image embedded watermark is acquired after IDWT. The experiment results show that the original image is not required in theprocedure of extracting the watermark when no attacking(NC = 1. 0000 ), the algorithm has good robustness on image cut, JPEG and median filter(NC〉0.5).
出处
《微电子学与计算机》
CSCD
北大核心
2008年第7期173-175,共3页
Microelectronics & Computer
基金
教育部科研基金重点项目(20030502)
河南省教育厅自然科学基金项目(20015200036)