The algorithm divides the original image into blocks, makes two layers of wavelet decomposition for each block, makes a certain number of Arnold transformations to the watermark image, and takes the number as the key. In the embedding process, all the image blocks are divided into two groups, select the embedded location on each piece, and embed the same watermark in two groups of image block respectively. Watermark extraction: according to the embedding rules, extract watermark images in the corresponding position, make a certain number of Arnold transformations according to the key, and synthesize the two watermark images extracted into a more accurate image. The algorithm can resist common image processing methods and has strong robustness.%该算法对原始图像进行分块，对每一块进行两层小波分解，对水印图像进行一定次数的Arnold变换，这个次数作为密钥；在嵌入过程中，将所有的图像块分成两组，并且在每一块上选择嵌入位置，分别在两组图像块中嵌入同一水印。水印的提取：根据嵌入的规则，在相应位置提取两幅水印图像，再根据密钥进行一定次数的Arnold变换，将提取出的两幅水印图像合成为一个更加准确的水印图像。该算法能够较好的抵抗常见的图像处理方法，具有较强的鲁棒性。
展开▼
