摘要
为了渲染含大量动态光源的场景,提出一种层次重采样算法.首先提出一种稀疏八叉树的实时重建策略,通过并行排序和合并加速其重建;然后通过自顶向下遍历稀疏八叉树,对以均匀网格离散化表示的空间进行高效采样;最后,为了提高渲染质量,对该空间内的均匀网格进行重采样.在ORCA和自建场景下的实验结果表明,所提算法能够较好地在渲染效率和质量间取得平衡,采样和渲染质量优于随机光源裁剪算法,而且渲染结果质量相近时,所提算法采样开销可减少约50%.
A hierarchical resampling algorithm is presented for rendering scenes with many dynamic lights.Firstly,a strategy for sparse octree parallel reconstruction is proposed,which accelerates reconstruction by parallel sorting and merging.Then,with top-down sparse octree traversal,the space is efficiently sampled which discretely represented by uniform grids.Finally,rendering quality is improved by resampling grids in sampled space.The experimental results in ORCA and self-building scenes indicate that the proposed method can achieve a good balance between the rendering performance and quality.Comparing with real-time stochastic lightcuts algorithm,the proposed algorithm achieves better sampling and rendering quality.And when rendering results have similar quality,the sampling consumption is reduced by about 50%.
作者
陈健蔚
刘昊峰
刘艳丽
张严辞
Chen Jianwei;Liu Haofeng;Liu Yanli;Zhang Yanci(College of Computer Science,Sichuan University,Chengdu 610065;National Key Laboratory of Fundamental Science on Synthetic Vision,Sichuan University,Chengdu 610065)
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2022年第5期784-793,共10页
Journal of Computer-Aided Design & Computer Graphics
基金
国家重大项目(GJXM92579)。
关键词
多光源
层次采样
重要性采样
光线跟踪
实时渲染
many lights
hierarchical sampling
importance sampling
ray tracing
real-time rendering