Abstract

The reproduction of gloss on displays is generally not based on perception and as a consequence does not guarantee the best visualization of a real material. The reproduction is composed of four different steps: measurement, modeling, rendering, and display. The minimum number of measurements required to approximate a real material is unknown. The error metrics used to approximate measurements with analytical BRDF models are not based on perception, and the best visual approximation is not always obtained. Finally, the gloss perception difference between real objects and objects seen on displays has not sufficiently been studied and might be influencing the observer judgement.

This thesis proposes a systematic, scalable, and perceptually based workflow to represent real materials on displays. First, the gloss perception difference between real objects and objects seen on displays was studied. Second, the perceptual performance of the error metrics currently in use was evaluated. Third, a projection into a perceptual gloss space was defined, enabling the computation of a perceptual gloss distance measure. Fourth, the uniformity of the gloss space was improved by defining a new gloss difference equation. Finally, a systematic, scalable, and perceptually based workflow was defined using cost-effective instruments.

Library of Congress Subject Headings

Computer graphics--Quality control; Imaging systems--Image quality--Evaluation; Art--Reproduction--Evaluation; Reflectance--Measurement; Color vision

Publication Date

11-21-2014

Document Type

Dissertation

Student Type

Graduate

Degree Name

Color Science (Ph.D.)

Advisor

Mark D. Fairchild

Advisor/Committee Member

Reynold Bailey

Advisor/Committee Member

Joe Geigel

Comments

Physical copy available from RIT's Wallace Library at T385 .F67 2014

Campus

RIT – Main Campus

Plan Codes

CLRS-PHD

Share

COinS