Perceived Roughness of Material Surfaces

Roughness is an important perceptual characteristic of material surfaces. This work investigated a derived model for evaluating surface roughness in accordance with human visual perception. This was accomplished with the help of extensive psychophysical experiments. Surface textures resembling natural materials were generated with different roughness. Subjects rated the surface textures on 5-point Likert scales based on visual perception of surface roughness. We trained a general model for mapping features of surface images to perceptual scales of roughness. It can be observed that the roughness obtained in this way is in good agreement with that perceived by human. Thus, given a material surface, roughness can be estimated accurately.