用阴极射线管显示器研究辨色阈值Ⅱ:典型色差公式评价

• 作 者: 汪哲弘 徐海松
• 机 构: 浙江大学现代光学仪器国家重点实验室,浙江大学现代光学仪器国家重点实验室 杭州310027,杭州310027
• 出 处: 光学学报
• 发表日期: 2007年07期
• 关键词: 视觉与色彩 颜色辨别阈值 性能因子 色差公式 色差预测性能

利用由辨色阈值实验得到的CIELAB颜色空间中等明度a*-b*平面上的视觉数据,对四个典型色差公式CIELAB、CMC、CIE94和CIEDE2000在辨色阈值水平上的色差预测性能进行测试和评价,并采用性能因子PF/3来表示各色差公式的优劣。对于原始形式(kC=kH=1)的色差公式,CIEDE2000的色差预测性能最好,其次为CIE94和CMC,CIELAB最差。采用对kC和kH进行优化后的色差公式,CIEDE2000的色差预测性能依然最佳,其余三个色差公式的预测性能基本相同;除了CIELAB明显改善外,其他公式的色差预测性能只比原始形式略有提高。色差公式CIEDE2000在黄色和蓝色区域的色差预测性能表现优异,但在灰色、红色和绿色区域的预测性能仍有待进一步改进。

The color difference prediction performance of the four typical color difference formulas CIELAB, CMC, CIE94 and CIEDE2000 at the color discrimination threshold level was tested by using the visual data in the medium lightness a * -b * plane of the CIELAB color space experiment obtained by the color discrimination threshold experiment and Evaluation, and the use of performance factor PF / 3 to indicate the merits of the color difference formula. For the color difference formula of the original form (kC = kH = 1), CIEDE2000 has the best color difference prediction performance, followed by CIE94 and CMC, and CIELAB is the worst. Using the color difference formula optimized for kC and kH, the color difference prediction performance of CIEDE2000 is still the best, and the prediction performance of the other three color difference formulas are basically the same. Except for CIELAB, the color difference prediction performance of other formulas is only slightly smaller than the original form improve. The color difference formula CIEDE2000 performs excellent color difference prediction performance in the yellow and blue areas, but the prediction performance in the gray, red and green areas remains to be further improved.