David C. Day

Abstract

Current imaging practices are based on exploiting metamerism to record and reproduce images. As a result, the data obtained in these images are dependent on the viewing conditions and the observer. While these methods produce acceptable results for day to day use, they often do not exhibit the type of accuracy and control required for scientific purposes such as art conservation. As a solution, many research institutions are now advocating the use of multispectral imaging to record the objects fundamental spectral properties to remove the data's dependency on the observer and viewing environment. The research described in this thesis involved determining if a trichromatic camera and readily available filters can be used for spectral estimation purposes. The Pixel Physics TerraPix camera system was characterized, its response to a target and 105 Kodak Wratten Filters under tungsten illumination was simulated, and spectral reflectance estimations were generated. The top filter candidates were chosen based on their simulated performance. These filters were then used in an imaging experiment designed to approximate conditions that would be found in an art gallery or other place where copy work is performed. The results of the imaging experiment were compared with the simulation, and shortcomings of the model were identified. The results of the experiment show that a camera model can be used as a guiding tool to make filter selections for spectral estimation.

Library of Congress Subject Headings

Spectral reflectance; Photography--Light filters--Testing; Digital cameras; Colorimetry

Publication Date

2003

Document Type

Thesis

Student Type

Graduate

Degree Name

Imaging Science (MS)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Roy Berns

Advisor/Committee Member

Francisco Imai

Advisor/Committee Member

Mark Fairchild

Comments

Physical copy available from RIT's Wallace Library at QC425 .D39 2003

Campus

RIT – Main Campus

Download

Share

COinS