PREDICTING THE IMAGE QUALITY OF COLOR OVERHEAD TRANSPARENCIES USING ACOLOR-APPEARANCE MODEL

Citation
T. Tanaka et al., PREDICTING THE IMAGE QUALITY OF COLOR OVERHEAD TRANSPARENCIES USING ACOLOR-APPEARANCE MODEL, Journal of electronic imaging, 6(2), 1997, pp. 154-165
Citations number
15
ISSN journal
10179909
Volume
6
Issue
2
Year of publication
1997
Pages
154 - 165
Database
ISI
SICI code
1017-9909(1997)6:2<154:PTIQOC>2.0.ZU;2-J
Abstract
The color-image quality of color overhead transparencies depends on pr operties of the imaging system used to create the transparency and ill uminating and viewing conditions of the transparency such as the proje ctor's spectral power distribution, projector distance from the screen , and luminance ambient lighting, screen gonio-spectral reflectance fa ctor, and viewing distance and geometry. As different visual fields an d/or luminance of those fields, some of these illuminating and Viewing conditions can be taken into suitable account using color-appearance models. A visual experiment is performed to determine whether color-ap pearance correlates of visual perception could be used to predict colo r-image quality for this imaging modality. The Hunt 1991 color-appeara nce model is used to define correlates of hue, brightness, colorfulnes s, lightness, and chroma for both pictorial and business-graphic scene s viewed under several combinations of ambient illuminance and project or luminance. Gamut volume is defined based on either absolute attribu tes-hue, brightness, and colorfulness-or relative attributes-hue, ligh tness, and chroma. Seventeen observers performed a preference experime nt generating interval scales of color-image quality. ii is found that gamut volume defined by using correlates of hue, brightness, and colo rfulness well predicted color-image quality. Of these correlates, colo rfulness was the most important factor. (C) 1997 SPIE and IS&T. [S1017 -9909(97)00302-4].