Improving the Perceptual Uniformity of a Gloss Space

The perceptual gloss space defined in Pellacini et al. [9] could be used for quality control applications to bring similar benefits as seen in color with the use of CIELAB. However, a distance metric to relate all the dimensions in the space does not exist, and the space was only validated with the materials used to define the space. The current space’s distance metric does not allow relating differences in lightness to the other dimensions: contrast gloss and distinctness of image gloss. The lightness perception uniformity of the space was first evaluated in a psychophysical study, where the observers’ lightness discrimination was found to decrease as lightness increased. A function was derived to model the lightness perception observed and it was included into the distance metric of the space. The space uniformity around sixteen positions in the gloss space was evaluated in a second psychophysical study to assess the overall space uniformity. The space was found to be perceptually non-uniform outside the samples used when the space was created. Also, an improved gloss difference equation that takes into account the non-uniformity of the space is presented, showing a statistical significant improvement over the current gloss difference equation of the space and reducing the STRESS value from 39.76 to 22.96. Introduction The CIELAB color space and the color difference equations defined in it have been widely used for quality control applications and to characterize input and output devices taking into account human color perception. When considering gloss perception, perceptual attributes defined as ASTM features (distictness-of-image gloss, haze, bloom, etc.) are used for quality control applications, and perceptual gloss spaces are discussed in the computer graphics literature. The perceptual gloss space defined in Pellacini et al. [9] could be used for quality control applications to bring similar benefits as seen in color with the use of CIELAB. For example, the ability to set tolerances across multiple dimensions and the opportunity to have a single value to describe the perceptual distance between two materials. However, some limitations still restrict the use of this perceptual gloss space in that regard. The first limitation is the need to locate the position in the space for a given real material, which could be accomplished by approximating material measurements with the underlying analytical model that defines the perceptual gloss space. The second limitation is the need to further characterize the relation between different dimensions in the space and to validate the space’s perceptual uniformity. This paper is going to focus on the uniform gloss space defined in Pellacini et al. [9], by first adding lightness to the gloss difference equation of the space, then the space uniformity will be evaluated, and finally an improved gloss distance measure will be presented. The current space’s distance metric does not allow relating differences in lightness to the other dimensions: contrast gloss and distinctness of image gloss. The lightness perception uniformity of the space was first evaluated in a psychophysical study, where the observers’ lightness discrimination was found to decrease as lightness increased. Then, a function was derived to model the lightness perception observed and it was included in the distance metric of the space defined in Pellacini et al. [9]. It’s important to note that the gloss perception using a common LDR display is studied in this work, and similar experiments would need to be performed to study the gloss perception of physical objects or when using HDR displays. The perceptual uniformity around sixteen positions in the gloss space was evaluated in a psychophysical study. The results of our study show that the space is not perceptually uniform and improvements on the distance metric are needed to improve its uniformity outside the samples used to create the space. Finally, the results of the second experiment are used to derive an improved gloss distance measure of the space. To summarize, the main contributions of this paper are: • The inclusion of lightness to the gloss difference equation of the gloss space defined in Pellacini et al. [9]. • The validation of the gloss space uniformity, finding that the space is non-uniform outside the samples used to create the space in Pellacini et al [9]. • The creation of an improved gloss difference equation that takes into account the space non-uniformity. Background In Pellacini et al. [9], a perceptual space of glossy materials represented by the Ward BRDF model [15] was presented. Two different experiments were performed to find the dimensionality of the gloss space for a set of samples, and then to perceptually scale the axes to obtain a uniform space. Two dimensions were found to be enough to describe the gloss perception: contrast gloss (c), which describes the relation between the diffuse component and the specular peak, and distinctness of image gloss (d), which defines the sharpness of the reflections. Lightness (L∗) was added to define the diffuse component, and the mapping from the Ward BRDF model parameters to the perceptual parameters was defined as: c = 3 √ ρs +ρd/2− 3 √ ρd/2 (1)