三维视频系统由于能够提供立体感、视点交互性的全新视觉体验而越来越受到人们的欢迎，被认为是下一代视频系统的主要发展方向。本项目研究解决两个基本的科学问题：一是如何建立各种感知特征对三维视觉的影响机理进行视觉质量计算模型和视觉舒适度计算模型的设计；二是如何将视觉质量计算模型和视觉舒适度计算模型进行融合，解决三维视频系统中的客观质量评价问题。主要研究内容为：通过分析影响三维视觉的感知特征，研究感知特征对三维视觉的影响机理，为质量评价提供理论依据；然后，将影响视觉质量和视觉舒适度的各种因素进行量化，建立视觉质量计算模型和视觉舒适度计算模型；最后，分别采用特征训练和模型调制两种方法，将视觉质量计算模型和视觉舒适度计算模型融合到客观评价方法中，使得质量评价能够更加准确地反映人类视觉感知。通过本项目的研究，将形成一套有效的三维视频客观质量评价的理论与方法，为三维视频系统的实现提供理论支持。

Three-dimensional (3D) video system has drawn increasing attention as it can provide 3D visual experiences of depth perception and viewpoint interaction, which is expected to be the main development direction for the next-generation video system. This project solves two fundamental scientific issues: The first issue is how to establish the mechanism of how the perceptual characteristics affect the 3D vision, so as to provide the theoretical basis for the design of visual quality and visual comfort calculation models. The second issue is how to solve the object quality assessment problem in 3D video by combining with the visual quality and visual comfort calculation models. The main research contents are as follows: Firstly, by analyzing the perceptual characteristic that affect 3D vision, the effect mechanism is researched, which provides the theoretical basis for quality assessment. Then, by quantifying a variety of factors that affect the visual quality and visual comfort, the visual quality and visual comfort calculation models are established. Finally, by adopting feature training and model modulation, the above visual quality and visual comfort calculation models are fused into the objective assessment method, which can accurately reflect human visual perception. Through this project, it will form more effective theories and methods for 3D video quality assessment, and provide further theoretical support for 3D video system.