Color Image Watermarking Scheme Based on Support Vector Regression

A color image watermarking scheme based on support vector regression (SVR) is proposed in this work. Considering the characteristic of human visual system (HVS), the watermark information is embedded in the luminance channel of YCbCr color space. The luminance component of original color image is partitioned into contiguous blocks. Some blocks with small variance values are selected to train SVR to learn the relationship between the center pixel and its neighborhood. Then, the watermark is embedded and extracted by comparing the luminance value and the output of the trained SVR. Experimental results show that the proposed scheme has good imperceptibility and robustness to common image processing.