光谱颜色管理系统关键理论和方法研究

The requirement of color image accurately cross-media reproduction is increasing in practice with the comprehensive, systematic and in-depth development of digital information technologies. The traditionally colorimetric representing and reproduction method based on metamerism could not meet various application requirements as the problems of color information losses and reproduction accuracy affected by lighting and observing environments. The key theories and methods of the spectral color management system, which could exactly represent and reproduce the color information, would be studied to implement the cross-media spectral reproduction of color images. Firstly, the spectral interim connection space with definitively colorimetric meanings and high practical values would be defined based on the analyzed causes and effects of the metamerism in the colorimetric reproduction system, and then the spectral gamut description, spectral gamut mapping algorithm, spectral color separation of the multicolorant printer would be implemented in the defined spectral interim connection space. Finally, the application examples of the digital conservation and cross-media reproduction of multispectral images would be implemented. The study results could be widely used in various fields with high cross-media reproduction accuracy requirements, such as the digital conservation and reproduction of the cultural artwork, digital publishing and distribution, telemedicine, e-commerce, and so on. The study results also have high theoretical reference values for the primary color selection and spectrally separating algorithm design in the developing process of multicolorant devices.

随着数字信息技术的全面系统纵深化发展，实际应用中对彩色图像跨媒体复现的颜色精度要求越来越高。传统基于同色异谱的颜色表征及色度复现方法存在颜色信息丢失及彩色图像跨媒体复现精度受照明观察环境影响等问题，难以有效满足各种应用要求。本项目对能够完整描述及复现颜色信息的光谱颜色管理系统中的关键理论和方法进行研究，实现彩色图像跨媒体准确的光谱复现。首先对色度复现系统中同色异谱的成因及影响进行分析，在此基础上定义具有明确颜色意义且实用价值较高的光谱连接空间，然后在光谱连接空间中实现设备光谱色域描述、色域映射算法设计以及多基色打印机光谱分色，最后给出多光谱图像数字典藏及跨媒体复现应用实例。研究成果可以广泛用于文化艺术品的数字典藏及复制、数字出版及发行、远程医疗、电子商务等众多对彩色图像跨媒体复现精度有较高需求的应用领域，另外对多基色设备开发过程中的基色选取、分色算法设计等也有较高的理论参考价值。

本项目旨在解决彩色图像跨媒体复现过程中的颜色失真问题。在过去三年的研究期间，本研究团队始终围绕课题任务书中的研究目标，分别对申请书中的各项内容进行了研究，取得了一系列理论与实践研究成果，达到了预期研究目标，超额完成了预期成果指标。本研究团队首先针对现有ICC颜色管理系统用于从数字图像获取到显示时存在的颜色复现精度较低问题，设计相应模型算法对现有的ICC颜色管理系统进行了改进，并应用于网络颜色管理。另外针对ICC颜色管理系统存在的同色异谱问题，提出了光谱颜色管理系统框架，并对该框架中的关键核心技术进行了研究。搜集了25303486个自然界物体光谱并对其同色异谱程度进行统计分析，获得了同色异谱程度随光源及颜色区域变化的规律。提出了两种方法用于实现基于数码相机的多光谱图像获取。针对CIEXYZ三刺激值在光谱颜色信息描述上的不足，在CIEXYZ标准色度观察者基础上提出了l、m、s三维基向量，定义了XYZLMS六维光谱设备无关颜色连接空间。在XYZLMS光谱设备无关颜色连接空间中实现了物体光谱数据集及打印机的光谱色域边界描述，并在XYZLMS光谱设备无关颜色连接空间中建立了查找表，实现了光谱到打印机的色域映射及光谱分色。该项目研究成果可以广泛用于打印印刷、印染、文化艺术品的数字典藏及复制、远程医疗和电子商务等众多对彩色图像跨媒体复现精度有较高需求的应用领域。