宽色域显示设备的高质量图像复现关键技术研究

Excellent performances on image reproduction are essential to displays. Recently, wide color gamut displays with the capability to exhibit more vivid colors have become popular and widely used. However, when images from the conventional image or broadcasting standards are shown on wide gamut displays without additional processing, the mismatches between the device gamut and the standard gamut may induce in some color distortions. In order to solve the problem of depressed image reproduction quality caused by the gamut mismatches for wide color gamut displays, an in-depth study on relevant technologies will be conducted, aiming to make innovations on the gamut mapping methods and colorimetric characterization models by combing the image quality theory with the color reproduction procedure. Based on the influences of naturalness and colorfulness on image quality, the descriptions of mapping requirement would be proposed according to different image contents with their color reproduction intents respectively. Afterwards, the gamut mapping methods for the preferred color reproduction intent would be deeply investigated, in which the color gamut boundary description methods would be proposed based on the CIECAM02 color space, as well as the extension algorithms to fulfill the descriptions of mapping requirement for different image contents. Also, accurate and practical colorimetric characterization models would be established to show the results of image reproduction on wide color gamut displays. The study could solve this problem of the contradiction between vivid visual effects and realistic feelings for images shown on wide color gamut displays, and it would helpful to bring into play their advantages of wide color gamut capability, which would promote the developments and applications of wide color gamut displays.

高质量的图像复现对显示设备至关重要。目前，具备更广颜色呈现范围的宽色域显示设备已普及应用，但现行数字图像或电视标准的色域与之不匹配，不加处理直接显示可能造成图像复现时有些颜色失真。针对色域不匹配引起宽色域显示设备图像复现质量下降的问题，本项目拟对此展开深入研究，将图像质量理论结合于颜色复现流程，对色域映射和颜色特征化进行方法和关键技术上的突破。本项目基于自然性、鲜艳性等感知参量对图像质量的影响，针对不同图像内容提出在其相应复现目标下的映射需求表征，重点探索喜好复现目标下的色域映射方法，在CIECAM02色貌空间完成色域边界描述，根据不同复现目标的映射需求表征建立新扩展映射算法，并以精确实用的颜色特征化模型在宽色域显示设备上予以图像呈现。本项目能够解决宽色域显示时绚丽效果与真实感存在矛盾的问题，使宽色域显示设备颜色呈现能力的优势得以充分发挥，从而促进其进一步发展及推广应用。

本项目旨在通过理论方法和关键技术的突破，解决宽色域显示设备与现行图像标准之间色域不匹配导致图像复现质量下降的问题，从而满足宽色域显示设备更高图像质量和更优视觉效果的迫切要求。在本项目三年研究期间，研究团队始终围绕项目任务书中的研究目标，分别对申请书中的各项内容进行了重点研究，取得了一系列理论与实践成果，达到了预期目标。.主要研究工作包括：首先，根据颜色特征化理论，通过对不同宽色域显示设备的颜色特性测量及分析，均建立起其精确适用的正向、反向颜色特征化模型，完成三刺激值与数字驱动值间的双向转换，其中对于严重不满足通道独立性的显示设备，设计多项式补偿模型实现精确的正向预测；接着，选取自然场景和鲜艳场景源图像，采用彩度调制生成测试图像集，实施心理物理学实验，通过大量视觉数据的分析，考虑自然性、鲜艳性等感知参量对图像质量的作用，对自然场景图像、鲜艳场景图像及保真应用场景图像三类图像分别建立了映射需求表征的数学描述，即自然场景图像和鲜艳场景图像依据喜好复现目标，保真应用图像依据等价复现（或对应复现）目标；继而，针对两种复现目标，在CIECAM02色貌空间中相应设计了映射流程及其算法实现，实现高效、准确的色域边界描述，完成色域映射并获取复现图像；最后，在三台宽色域显示设备上呈现图像复现效果，以心理物理学实验视觉评估映射算法的实用性能，结果表明本项目提出的映射方法可以获取更优图像复现效果；此外，针对实际应用中宽色域显示设备的最大色域设置并非最优显示状态这一问题，设计实现了一种显示设备的三原色优化选取方案及实现流程，提出了颜色喜好性指数获取显示设备的最优三原色设置方案，使其宽色域特性在更合理的参数设置下展示。.综上，本项目通过一系列关键技术实现了宽色域显示设备的高质量图像复现，使宽色域显示设备的颜色呈现能力这一优势可在应用中充分发挥。项目成果对显示行业具有积极意义，有利于宽色域显示设备的持续发展和更好推广。