Human visual system-based perceptual Mura index for quantitative Mura evaluation

We propose a new quantitative Mura evaluation metric that refers to a human perceptual Mura index (HPMI) for a given captured panel image including a Mura artifact, which considers the perceptual differences of Mura features based on the human visual system (HVS). Conventional quantitative Mura evaluation metrics are highly dependent on the contrast feature of the Mura region, in which perceptual Mura level can vary depending on the perceptual characteristics with background gray levels (BGLs) in addition to the contrast. Although various studies have tried to solve the intrinsic weakness of a contrast-based metric caused by insufficient treatment of perceptual Mura features, there is still room for reflecting the variations of human perception caused by BGLs and Mura types with HVS properties. To solve this problem, we provide two solutions to evaluate the Mura level that can reflect the perception characteristics of human eyes. First, we establish the individual evaluation metrics depending on the BGLs by formulating the relationship between the human inspection and Mura level based on the perceptive features in the Mura region. Second, we apply adaptive HVS-based preprocessing to the contrast map of the Mura image, which represents the different ratios of variation in the Mura region and background region depending on the Mura types. Consequently, the correlation between subjective ranking by multiple human inspectors and objective ranking by the proposed HPMI increases considerably, up to 0.559 at the low BGL, compared with that of benchmark methods. Furthermore, by applying HVS-based preprocessing, the correlation for subjective ranking is improved up to 0.77 in line Mura.