Assessment of display quality is important for display manufacturers. This should be evaluated with two aspects – measuring physical performance and quantifying human perception of image appearance attributes. The present work particularly focuses on the development of an image quality model for the latter case. A developed model included three components (colorfulness, contrast and naturalness) that were all modeled using parameters derived from CAM02-UCS color-appearance attributes. In the combination of the three components, psychophysical relationships between each of them and image quality were exploited. The results exhibited that colorfulness and naturalness were critical factors in the judgment of image quality. A computational procedure was also given from a pair of color images to predict the image quality.