Premium
Colorimetric characterization of a computer‐controlled liquid crystal display
Author(s) -
Day Ellen A.,
Taplin Lawrence,
Berns Roy S.
Publication year - 2004
Publication title -
color research and application
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.393
H-Index - 62
eISSN - 1520-6378
pISSN - 0361-2317
DOI - 10.1002/col.20046
Subject(s) - colorimetry , liquid crystal display , gamut , computer science , optics , artificial intelligence , materials science , computer vision , physics , operating system
Abstract A new method was used to characterize computer‐controlled liquid crystal displays (LCDs). The characterization, which was performed to enable colorimetric image display, included channel independence, spatial independence, screen uniformity, and colorimetry. The colorimetric model consisted of three one‐dimensional look‐up tables (LUTs) describing each channel's optoelectronic transfer function and a 3 × 4 matrix transformation that included black‐level flare. The matrix coefficients were estimated statistically by minimizing the average CIEDE2000 color difference for a data set sampling the display's colorimetric gamut. The LUTs were recreated dynamically throughout the optimization of the matrix coefficients. The characterization was implemented with three different instruments to evaluate the robustness of the method with respect to measurement uncertainty. The average performance ranged between 0.1 and 0.4 ΔE 00 and was well correlated with instrument precision. The optimization approach improved performance by a factor of two compared with direct measurements. Despite differences in instrument design, the chromaticities of each primary following optimization and black‐level flare compensation were very similar. This excellent performance was a result of the display's optoelectronic properties well matching the model assumptions. The technique was also used to characterize three additional LCD displays ranging in their matching of the model assumptions. In this case, performance worsened. For one display, more complex models would be required for colorimetric characterization. Finally, a colorimetric characterization based on measurements at the center of the display and perpendicular to the face was used to predict measurements at the edges and at different angles. The results indicated that characterizations would be required at multiple positions and angles in order to achieve sufficient accuracy. © 2004 Wiley Periodicals, Inc. Col Res Appl, 29, 365–373, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.20046