HSV颜色空间的彩色图像数字水印算法

A WATERMARKING ALGORITHM FOR COLOR IMAGES BASED ON HSV COLOR SPACE

提出了一种新的基于HSV颜色空间的彩色图像数字水印算法。该算法采用彩色图像作为水印嵌入到原始彩色图像中。算法根据HSV颜色空间的特点,首先将RGB格式的彩色水印图像转换到HSV颜色空间内,然后分别将水印图像的H、S、V三层进行分块DCT变换,对每一个子块保留左上角的系数,将其它的系数替换为0,得到嵌入矩阵,使得嵌入的有效信息大为减少。再将原始彩色图像转换到HSV域内,对原始图像的H、S和V平面进行分块DCT变换后,将变换后的H、S、V层水印信息分别嵌入到原始图像的H、S和V层变换后的中频和低频系数中,最后通过DCT逆变换和颜色空间转换,将嵌入水印的图像转换成RGB的图像,得到嵌入水印的图像,完成水印的嵌入过程。水印图像小于载体图像,利用冗余嵌入提高了水印的鲁棒性。

Based on HSV（ hue, saturation, and value） color space, it proposes a watermarking algorithm for color images. The digital watermark using the color image is embeded in the original color image. According to the characters of HSV color space, and using a different color space, a RGB image is processed by first converting it to an HSV image. And then transform the watermarking images of H plane, S plane and V plane while using DCT by 8 ＊ 8 blocks. Keep DCT coefficients in the top left comer of every block and replace other coefficients with zeros in order to reform new matrix with fewer effective coefficients. Then, convert the original color image to an HSV image. Embed watermarking image in intermediate and low frequency of coefficients DCT after transforming the H, S and V planes using DCT by 8 ＊ 8 blocks. Then a new watermarked image forms through the inverse DCT transform. At last,convert it back to RGB color space. Watermarking image is smaller than the original image,and the robustness is enhanced using redundant embedding technique.