Gamut mapping is a way of mapping color information in a source device or image to a target device or image.Because the color mechanisms of different devices are different,the loss of color information is inevitable during the process of gamut mapping.So there are not only gray distortions but also serious color distortions in gamut-mapped images(GMIs).So far,the image quality measures of GMIs extract the features that can represent gray and color distortions from gray domain and color domain respectively.But the gray value of the image is obtained by linear calculation of three color channels(e.g.R,G and B),and the change of the gray information is reflected in the three color channels.Based on this,the assumption was proposed that R,G and B color information can represent not only color distortion but also gray distortion of image in this study.On the basis of this assumption,quality perceived features were extracted from R,G and B color components directly,and the image quality evaluation model was trained based on the back propagation neural network.Finally,experiments on three public GMI databases prove the superiority of this algorithm.
Gamut-mapped imagesImage quality evaluationColor distortionGray distortionQuality perceived features
NETL
NSTL
万方数据
