Purchase this article with an account.
Barton L. Anderson, Phillip J. Marlow, Juno Kim; Disentangling 3D Shape and Perceived Gloss. Journal of Vision 2012;12(9):947. doi: 10.1167/12.9.947.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent work has shown that judgments of gloss are distorted by 3D surface geometry. These results reveal consistent failures in gloss constancy, but the cause of these failures remains unknown. We conducted experiments to determine what underlies the effects of 3D surface relief on perceived gloss. In Experiment 1, bumpy surfaces with a fixed gloss level and different levels of relief were rendered in a number of different illumination fields that varied in complexity and direction of the primary light source(s). Surfaces were viewed either with or without binocular disparity. Observers performed paired comparisons of perceived gloss for all stimuli within a given illumination field. The results of these experiments revealed that gloss was strongly affected by 3D surface relief, but that the effects of relief varied significantly across different illumination fields and disparity conditions. Perceived gloss either increased monotonically, decreased monotonicaly, or exhibited strong non-monotonicities as a function of relief height. We hypothesized that all of these effects arose from a number of simple image cues that co-varied with observers' gloss judgments: the contrast, disparity, coverage, and sharpness of specular reflections. To test this hypothesis, we performed an experiment in which independent groups of observers performed paired comparisons and judged each image attribute separately (specular contrast, disparity, coverage, and sharpness). A separate group of observers judged gloss for the same stimuli. Observer’s gloss judgments are very well fit by a simple weighted average of the data obtained by observers’ judgments of specular contrast, coverage, sharpness, and disparity. Our results suggest that the "contamination" of gloss judgments by 3D shape is caused by the effects shape has on simple image properties that modulate the perception of gloss. Our results suggest a general psychophysical method for testing the role of image cues to material properties.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only