Estimating the glossiness transfer function induced by illumination change and testing its transitivity
Katja Doerschner1, Huseyin Boyaci, Laurence T Maloney
1Department of Psychology, Bilkent University, Ankara, Turkey. katja@bilkent.edu.tr
Journal of Vision
|May 15, 2010
Summary
Perceived glossiness changes with lighting conditions. Researchers found that the relationship between glossiness and lighting can often be predicted linearly and transitively, though individual variations exist.
Area of Science:
- Visual perception
- Computer graphics
- Material science
Background:
- Surface glossiness perception is influenced by object reflectance and scene lighting (light field).
- Understanding how perceived glossiness changes across different light fields is crucial for realistic rendering and visual science.
Purpose of the Study:
- To determine the mathematical form of the glossiness transfer function between different real-world light fields.
- To test the transitivity of glossiness perception: if glossiness transfer from light field 1 to 2 and 2 to 3 is known, can it predict transfer from 1 to 3?
Main Methods:
- Four observers compared the perceived glossiness of surfaces under two distinct real-world light fields.
- Data was used to estimate a transfer function modeling glossiness remapping between light fields.
- The transitivity property of these transfer functions was statistically evaluated across sets of three light fields.
Main Results:
- Observers' estimated glossiness transfer functions were best modeled by a linear function.
- Three out of four observers demonstrated the expected transitivity pattern in their glossiness perception.
- The deviation from transitivity for one observer was minimal, accounting for less than 12.5% of the gloss range.
Conclusions:
- The perceived glossiness transfer function between different light fields can be approximated by a linear model.
- Glossiness perception generally exhibits transitivity, suggesting predictable relationships across varying lighting conditions.
- Individual differences in glossiness perception exist but appear to be relatively small within the tested range.

