中文摘要：

基于同时亮度对比的用户认知特点，构建了用户亮度感知的定量模型，为计算机界面设计中的亮度设置提供定量参考依据．在计算机仿真的CIE标准照明体D65照明下，采用仿真Munsell颜色体作为实验刺激，呈现在经过BP神经网络标定的CRT显示器上，仿真实验刺激的色差远小于最小颜色可觉差．基于亮度匹配的实验范式，4名男性用户参与了4（实验表面）×4（匹配背景）的被试者内设计实验，对每种实验条件进行了16次重复测量．重复测量方差分析结果表明，实验表面亮度和匹配背景亮度均对用户的匹配亮度产生了显著影响．用背景亮度作为自变量，匹配亮度和匹配背景亮度的韦伯对比度作为因变量，采用幂函数形式对实验数据进行拟合，在各种实验条件下拟合模型的R方均大于0．99．额外采用5名男性用户进行同样的实验对所构建模型进行验证，在各种实验条件下拟合模型的R方均大于0．95．上述结果表明，拟合模型对用户感知亮度随实验表面亮度和匹配背景改变而变化的趋势进行了高精度描述．

英文摘要：

Based on the user cognitive features of simultaneous brightness contrast, a qualitative model of user perceptive brightness was constructed for setting the value of brightness during the computer interface design. Simulated Munsell color chips under simulated CIE standard illuminant D65 were adopted as the test stimulates, when presented on a BP neural network calibrated CRT monitor, its color difference was far less than the minimum perceptible color difference （MPCD）. With the test paradigm of brightness matching, a 4 （test surface） by 4 （matching background） by 16 times repeated measures within subjects design test was conducted on 9 adult male users, 4 users＇ data for model parameter estimation, another 5 users＇ data for model assessment. Repeated measures ANOVA results showed that the 4 users＇ matched brightness was significantly influenced by the brightnesses of test surface and matching background. The 4 users＇ test data were fitted by a series of power function, with the background brightness as independent variable, the Weber contrast which was made of the matching surface brightness and background brightness as dependent variable. The results showed that all the R squares of fitted models were larger than O. 99. Then the fitted model was evaluated by another 5 users＇ model assessment data, and all evaluated R squares were larger than 0.95. With a fantastic precision, the fitted model described how the user pereepted brightness varied with the brightness variations of the test surface and the test background.