真实感材质外观的表示和高效渲染技术研究现状及展望  

Current Status and Prospect of Research on Modeling and Rendering Techniques for Realistic Materials

作　　者：陶成之 孙奇 过洁[1] 袁军平 周辰熙 何雪滟 范之闽 史鹏程 郭延文[1] Tao Chengzhi;Sun Qi;Guo Jie;Yuan Junping;Zhou Chengxi;He Xueyan;Fan Zhimin;Shi Pengcheng;Guo Yanwen(State Key Laboratory for Novel Software Technology(Nanjing University),Nanjing 210033)

机构地区：[1]计算机软件新技术全国重点实验室(南京大学),南京210033

出　　处：《计算机辅助设计与图形学学报》2024年第8期1131-1154,共24页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金重点项目(62032011).

摘　　要：在虚拟场景的组成中,除了三维几何模型之外最重要的要素就是物体表面的材质外观,对材质进行基于物理的建模和真实感绘制是保证虚拟场景真实感的重要手段.然而,由于现实世界材质外观本身的多样性和复杂性,迄今为止对真实感材质的研究一直是计算机图形学领域的热点和难点.文中总结了材质外观的近期相关工作,从材质的表示建模和材质的真实感绘制2大方向展开讨论;进一步,将材质模型分为表达范围广的通用材质模型和针对毛发及布料等材质的特殊材质模型2大类;并将材质绘制方法分为用于离线渲染的点采样优化和用于实时渲染的材质预过滤2种;最后指出材质研究可能的3个发展方向:神经网络材质、波动光学材质和材质的统一标准,为未来该领域的研究提供思路.In the composition of virtual scenes,besides three-dimensional geometric models,the most im-portant element is the appearance of object surfaces.Physically-based modeling and realistic rendering of materials are crucial means to ensure the realism of virtual scenes.However,due to the diversity and complex-ity of real-world material appearance,research on realistic materials has been a hot and challenging topic in the field of computer graphics.In this paper,we summarize numerous relevant works on material appearance,categorizing them into two main aspects:appearance modeling and realistic rendering of materials.Further-more,material models are divided into two major categories:general material models with board representa-tional space and special material models tailored for materials such as hair and fabric.Material rendering methods are also classified into two types:point-sampling optimization for offline rendering and material pre-filtering for real-time rendering.Finally,three potential directions for material research are pointed out:neural network materials,wave optics materials and a unified standard of material,providing insights for future research in this field.

关 键 词：真实感绘制 材质 表面微结构 毛发渲染 皮肤渲染 布料渲染 重要性采样 纹理预过滤 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]