基于噪声模型的DCT域盲水印算法

针对DCT交流系数分布函数的广义高斯分布模型设计了一个盲水印算法,该算法独立于具体图像,具有通用性.实验结果显示,该算法表现出良好的检测性能和对包括JPEG压缩、裁剪、高斯噪声、虚假嵌入在内的多种攻击的鲁棒性;进而,将此算法推广到多重水印嵌入场合,提出了三种重复嵌入策略,增加了嵌入的信息量和嵌入方式的灵活性.

Blind watermarking algorithm in DCT domain based on the noise model

This paper presents an image digital watermarking algorithm in the block-wise DCT domain. The algorithm is designed based on the General Gaussian Model, the special case of which can be the Gaussian Model or the Laplacian Model. No original image is needed when the watermarking being detected. Moreover, the algorithm is independent of the image, which makes the algorithm generic. The experimental results show the algorithm can resist to the attacks such as JPEG compression, cropping, histogram equalization and so on. Moreover, the algorithm is extended with three strategies, where more watermarks can be embedded at different positions of the DCT ac coefficients at the same time. With these, the practicability of the algorithm is increased.