基于DWT与DCT的灰度水印图像嵌入算法

图像中嵌入的多为二值水印，容量有限且抗攻击能力有待提高。为了增加嵌入水印的容量以及增强算法的鲁棒性，提出基于DWT与DCT的灰度水印图像嵌入算法。首先对灰度水印进行Arnold置乱预处理，其次对所选取的载体图像进行二级小波变换，选择第二层的逼近子带LL2与细节子带LH2、HL2、HH2依次嵌入灰度水印的高四位位平面。嵌入水印时对4个子带分别进行8×8分块DCT变换，然后采用一种改进的交换中频系数的方法将置乱后水印的高四位位平面分别嵌入所选取的4个子带，为增强算法的鲁棒性，在不同的子带由实验选择不同的水印嵌入因子。仿真实验表明，本算法抵抗图像裁剪、图像加噪、直方图均衡化、常用滤波处理、JPEG压缩等攻击的效果较好。

An Embedded Algorithm for Gray Watermarking Image Based on DWT and DCT

For embedding in the cover image are mostly binary watermarks, capacity is limited and needs to be improved anti-at- tack capability. In order to increase the capacity and enhance watermarking robustness of the algorithm, this paper proposes an embedded algorithm for gray watermarking image based on DWT and DCT. Firstly, the gray watermark image is preprocessed by Arnold scrambling. Secondly, after decomposing the selected cover image into the secondary layer, LL2, HL2, LH2 and HH2 of the secondary layer are selected to embed watermark. We apply 8 ~ 8 block DCT to each band and adopt an improved approach to embed high four bit planes of the gray watermark by switching two middle frequency coefficients. In order to enhance the robust- ness, the embedding intensity varies with different subbands. Finally, the simulation experiments show that the method is of strong robustness to cutting, image adding noise, histogram equalization, filtering, JPEG compression, and other attacks.