FCS-based Prediction in Color Changes under Different Illuminants
Spectral imaging has advanced to reproduce the real colors under different illuminants. Since multi- spectral camera is expensive, the more simple ways to predict the color renditions by RGB camera have been proposed actively. Human vision is characterized by metamerism that pair of colors match under one illuminant but differ under another illuminant. The projection from spectral to tri-color space is many- to-one and non-invertible. The projector from spectral to Human Visual Sub-Space is mathematically described by J. B. Cohen in FCS (Fundamental Color Space). This paper discusses a FCS-based estimation of color changes under different illuminants by using only XYZ or RGB tri-color data. In FCS, a spectral color input is decomposed into two components, fundamental and metameric black. Since the fundamental carries the intrinsic XYZ sensation to human vision and is invertible by a pseudo-inverse projection from XYZ, first, the color change is roughly estimated by biasing the fundamental with SPD (Spectral Power Distribution) of a given illuminant. Though the metameric black is an independent variable and unpredictable, next, the paper proposes a parametric prediction method for metameric black component using a correlation analysis back to XYZ space from FCS. Experimental results are presented for IT8 standard color target or a set of inkjet color chips under the illuminant changes from D65 to D50 or D65 to tungsten lamp A.
human vision color rendition spectral estimation illuminant change
Hiroaki Kotera
Kotera Imaging Laboratory, Chiba, Japan
国际会议
杭州
英文
134-137
2007-07-12（万方平台首次上网日期，不代表论文的发表时间）