CIECAM02的若干问题及解决办法

发布时间：2018-01-07 16:19

本文关键词：CIECAM02的若干问题及解决办法　出处：《辽宁科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：自从2004年CIETC8-01推荐CIECAM02色貌模型作为色彩管理应用以来,它便被用来在一个很广的观测条件下来预测色貌,主要在感知属性方面来指定色貌,来量化色差,并且为颜色管理提供了一个统一的颜色空间和特征文件连接空间。然而经过大量应用研究,Brill和Süsstrunk发现了CIECAM02存在“黄-蓝问题”和“紫色问题”。本文中具体分析了引发“黄-蓝问题”和“紫色问题”的原因,并且得到了一个结论:如果能满足一个嵌套规则,则可以同时解决“黄-蓝问题”和“紫色问题”,该规则被定义为:“如果希望所有位于光谱轨迹内的颜色取样经转换后不出现问题,则需满足:HPE三角形必须包含(改进的)CAT02三角形,而(改进的)CAT02三角形必须包含光谱曲线。”本文使用最优化解决问题的方法得出了解决方案得到了一个新的CAT02矩阵作为最优解。新的矩阵不仅满足了嵌套关系,还发现该模型在预测相应颜色和色貌数据集时,预测效果比Hunt-Pointer-Estevez(HPE)矩阵性能优越。此外,文章还通过实验数据分析了原始CAT02矩阵M02,新的矩阵MOPT和HPE矩阵MHPE之间的性能,结果发现新的矩阵和HPE矩阵在所有相应色数据集(共有21组数据)中对9组数据的预测结果存在很大差异。但总体来说,新的矩阵MOPT还是不如原始CAT02矩阵M02的预测效果好(平均差1.0个CIELAB单元)。最后文章得出结论:在既解决“黄-蓝问题”和“紫色问题”的情况下,又要保持对相应色数据集和色貌数据集具有良好预测能力的情况下,在以后的CIECAM02版本中本文推荐使用矩阵MOPT。
[Abstract]:Since CIETC8-01 recommended CIECAM02 color feature model as a color management application in 2004, it has been used to predict color appearance in a wide range of observation conditions. Mainly in the sense of attributes to specify color appearance, to quantify color difference, and for color management to provide a unified color space and feature file connection space. However, after a lot of application research. Brill and S 眉 sstrunk discovered the existence of "yellow-blue problem" and "purple problem" in CIECAM02. The reason. And a conclusion is drawn that if a nested rule is satisfied, the "yellow-blue problem" and "purple problem" can be solved at the same time. The rule is defined as: "if it is desired that all color samples located in the spectral locus will not have a problem after conversion, then the 1: HPE triangle must be included (improved CAT02 triangle)." And (improved). The CAT02 triangle must contain a spectral curve. " In this paper, we obtain a new CAT02 matrix as the optimal solution by using the optimization method to solve the problem. The new matrix not only satisfies the nesting relation. It is also found that the prediction performance of the model is superior to that of the Hunt-Pointer-Estevezia HPE matrix in predicting the corresponding color and color feature data sets. The performance of the original CAT02 matrix M02, the new matrix MOPT and the HPE matrix MHPE is also analyzed by the experimental data. The results show that the new matrix and the HPE matrix have very different prediction results for 9 groups of data in all the corresponding color data sets (there are 21 sets of data), but generally speaking. The new matrix MOPT is still not as good as the original CAT02 matrix M02 (average difference is 1.0 CIELAB units). Finally, it is concluded that the "yellow-blue problem" and "purple problem" are solved. In order to maintain good prediction ability for the corresponding color data sets and color appearance data sets, this paper recommends the use of matrix MOPTT in future versions of CIECAM02.
【学位授予单位】：辽宁科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：O151.21

【相似文献】