基于小波变换和纹理特征的水印算法的研究

目前基于小波变换的数字水印算法主要根据频域小波系数的幅值大小确定是否嵌入水印，这种算法的缺点是嵌入水印后的图像的视觉透明度不够好。对目前已有的基于小波变换的数字水印算法进行了研究，在保证嵌入水印的鲁棒性的前提下，也保证嵌入水印后的图像的理想的视觉效果的基础上，提出了一种基于小波变换和图像纹理特征的自适应数字水印算法，该算法利用了图像本身固有的纹理特征来寻找水印的嵌入位置，实验结果表明该方法能较好地降低原始图像视觉失真和经压缩变换后水印图像的失真。

Study of watermarking algorithms based on wavelet transform and texture features

Most wavelet-based digital watermarking algorithms determine whether to embed the watermark according to the magnitude of the wavelet coefficients, but this kind of schemes does not provide good image visual appearance after watermarking. By analyzing the characteristics of wavelet-based watermarking algorithms to preserve both the robust property and the visual appearance before and after watermarking, an adaptive visible watermarking scheme based on texture features in wavelet domain is proposed, which use the texture features of the image to find the embedding watermarking position. The experimental results show that the source image distortion and the compressed watermark image distortion are reduced.