The method for mapping linear polarization imaging variables to the color channels of hue, saturation, and value in the HSV color space is a common technique for the visualization of polarization imaging data. This method utilizes the structural similarities between polarization vision and color vision so that the full linear polarization information is depicted in a single image. Recent developments have attempted to address issues, arising from the fact that the HSV color space is not an accurate model of human color vision, by mapping the polarization channels of intensity, degree, and angle to the color channels of lightness, colorfulness, and hue defined in the perceptually uniform color space CAM02-UCS. While the theoretical benefits of this method have been demonstrated using metrics of perceptual uniformity and channel independence, the practical benefits to human observers has not been studied. In this user study, the two methods are compared using a series of forced-choice questions on the perceived magnitude differences in DoLP value to determine 1) which method produces fewer errors, 2) which method produces a more linear scale in degree of polarization perception, 3) whether the perception of degree of polarization is independent of intensity and angle of polarization.
|