基于QR码与混沌加密的Contourlet域彩色图像盲水印算法

目的针对现行数字水印算法中抗几何攻击能力弱以及嵌入水印信息容量差的问题,提出一种Contourlet变换下QR码与混沌加密相结合的彩色图像水印算法。方法采用QR码作为水印图像,利用Logistic混沌序列对其进行加密,再将加密的水印图像嵌入到二级Contourlet变换的低通子带中,利用系数间的大小关系,修改系数完成水印嵌入与提取。结果通过实验结果可知,水印的嵌入对载体图像影响较小,PSNR值高达42.764,经过旋转、剪切及常规攻击后,提取出的水印均可正确解码。结论该算法具有良好的安全性、不可见性,对于各类攻击具有较优的鲁棒性。水印信息容量高于普通图像,水印提取方式为盲提取,能够更好地满足各类数字作品版权保护的实际需要。

Blind Watermarking Algorithm for Color Image in Contourlet Domain Based on QR Code and Chaotic Encryption

To put forward a method of color image watermarking algorithm based on the combination of QR code and chaotic encryption that under Contourlet transform to solve problems of weak geometric attack ability and poor capacity of embedded watermark information in current digital watermarking algorithm. QR code image was used as the watermark. The logistic chaotic sequence was adopted to encrypt the watermark image. Then the watermark image encrypted was embedded into the low pass subband of the secondary Contourlet transform. The relationship between coefficient sizes was used to modify the coefficient and to complete watermark embedding and extraction. Watermark embedding had few influences on the carrier image and the PSNR value could reach 42.764. After rotation, shear and conventional attack, all the watermark extracted can be decoded correctly. The algorithm not only has good security and invisibility but also has good robustness to all kinds of attacks. The information capacity of watermark is higher than that of the ordinary images. The watermark extraction method is blind extraction, which can better meet the practical needs in copyright protection of various digital works.