Color Image Quality Assessment Based on Quaternion Singular Value Decomposition

An effective assessment method for color image is proposed. It is based on the quaternion description for the structural information of color image. The local variance of the luminance layer of color image is taken as the real part of a quaternion, then the three RGB channels of the color image are encoded into the three imaginary parts of the quaternion. The angle between the singular value feature vectors of the quaternion matrices correspond to the reference image and the distorted image is used to measure the structural similarity of the two images. Results from experiments show that the proposed method is better consistent with the human visual characteristics than MSE, PSNR and SSIM. The images whose size is different from that of the reference image can also be assessed by this method.