根据人类视觉系统(HVS)特征和小波变换的多分辨分解特性,提出了一种基于小波变换的图像数字水印算法。该算法在原始图像的LL域自适应地嵌入盲水印。提取或检测时无需参考原始图像或原始水印,实现了全盲检测,与其他许多算法相比,提高了水印的强度,且兼顾了水印的容量,计算简单,易于实现。实验结果表明,该方法具有较强的鲁棒性和安全性,适用于数字图像的版权保护。
According to the characteristics of Human Visual System (HVS) and the property of multi-resolution of wavelet transform , present a kind of image digital watermarking algorithms based on wavelet transform. In the LL subband of the original image we adaptively embed a blind watermarking. When the watermarking is extracted or checked, we need not refer to the original image or original watermarking. In this way, we bring totally blind check into realization. Compared with other algorithms raised the strength of the watermark, and give attention to the capacity of the watermark. The algorithms computes in brief, is easy to actualize. The result of the experiment shows that this scheme has higher quality of robustness and security against attacks and is applicable to the copyright protection of the digital picture.