Study On Color Inverse Halftoning Image Quality Assessment Based On Human Visual Properties

The traditional objective image assessment methods cannot reflect the visual perception of human well, so taking into account the process human perceives visual information is a beneficial way for image assessment. Study image quality assessment on color inverse halftoning based on human perception of color difference and habit human receive visual information, and two assessment methods are proposed.The color difference calculated by CIE 1976 La*b* formula doesn't match the characteristics of Human Visual System when the two colors are close especially. An improved quality assessment algorithm on inverse halftoning image is proposed, in which CIE DE2000 formula and the characteristics of Human Visual System have been applicated. Most image assessment methods based on human visual properties only take into account the early stage of the HVS. But natural image signals are highly structured, and the human visual system is highly adapted to extract structural information from the viewing field. Based on the two points above, an assessment method with structural similarity is proposed. The experimental results indicate that the two methods correlate with Mean Opinion Score well.