Humans are constantly making judgments about the world based on material appearance (e.g. is something old or new, clean or dirty, soft or hard?). A material appearance editor is an essential feature in any software system for generating synthetic photorealistic images. Increasingly the applications that rely on photorealistic imagery -- design, historical reconstruction, training, and entertainment -- depend on renderings of high visual quality. Ultimately human perception determines what is required for high visual fidelity. However, very little work has been done modeling human perception of materials in a form that can be applied to computer graphics rendering systems. In this project, the researchers are conducting a series of physical and psychophysical experiments with the goal of developing a material appearance model that is applicable to computer graphics. The focus is a class of materials with clear economic impact -- materials used in architectural design. This work holds promise of a strong impact on architectural-design and material manufacture by providing high-quality visualization of architectural materials. It will accelerate developing architectural materials by allowing researchers to simulate material appearance with physical accuracy as well as perceptual fidelity. Additionally, the psychophysical experiments will provide new insights into human vision.

The goal of this work is to produce a material appearance model that informs the efficient modeling, measurement and representation of materials for visual simulation. The results obtained in the process of developing this model include: - an apparatus particularly suited to measuring architectural material appearance properties. - psychophysical measures of the perception of materials accounting for directional, scale, and surround effects that are typical of architectural applications. - simplified material models, based on applying psychophysical measures to physical data. - experimental results verifying the visual accuracy of simplified material models. - a common database for both physical and psychophysical measures of material appearance for architectural applications.

Agency
National Science Foundation (NSF)
Institute
Division of Information and Intelligent Systems (IIS)
Application #
1218515
Program Officer
Ephraim P. Glinert
Project Start
Project End
Budget Start
2012-09-01
Budget End
2016-08-31
Support Year
Fiscal Year
2012
Total Cost
$500,000
Indirect Cost
Name
Yale University
Department
Type
DUNS #
City
New Haven
State
CT
Country
United States
Zip Code
06520