一种自适应图像灰度水印算法

提出了一种改进的SPIHT压缩算法,用于灰度水印的预处理,以得到易于嵌入的二进制数据.并将水印自适应地嵌入图像的低频小波系数,使得嵌入水印能量不超过图像可容纳的噪声上限.实验表明,算法能增强水印的鲁棒性,同时不会明显降低其视觉效果,且提取出的水印易于直观判断.

Abstract：

An improved SPIHT compression algorithm for the pretreatment of gray-level watermarks is presented to get the binary system's data which can be easily embedded. The watermark is self-adaptively embedded into the low frequency wavelet coefficient of image, so that the embedded watermark energy will not exceed the top limit of the acceptable noise of the image. Experiments demonstrate that the watermark's robustness has been enhanced, the conflict between invisibility and robustness has been mitigated, and the drawn watermark can be intuitively recognized.