Color Image Watermarking Using Support Vector Regression

This work concentrates on the problem of color image watermarking based on support vector regression (SVR). The original color image was partitioned into contiguous 3x3 blocks. Each block was converted from the RGB color space to the YCbCr color space and the watermark information was embedded in the luminance channel of YCbCr color space. Some blocks with small variance in luminance channel were selected for SVR which was trained to learn the relationship between the center pixel and its neighborhood. The watermark was embedded by comparing the luminance value and the output of the trained SVR. Experimental results show that the proposed method has good imperceptibility and robustness against common image processing.