Resampled importance sampled stochastic lightcuts

Konferenz: ICETIS 2022 - 7th International Conference on Electronic Technology and Information Science
21.01.2022 - 23.01.2022 in Harbin, China

Tagungsband: ICETIS 2022

Seiten: 4Sprache: EnglischTyp: PDF

Autoren:
Chen, Jianwei; Lin, Gang; Li, Zhengze (College of Computer Science, Sichuan University, Chengdu, China)

Inhalt:
Scenes with many dynamic lights exist in various industries, and rendering which with ray tracing in real time remains a challenging problem. In this paper, we extend the real-time stochastic lightcuts method to achieve better performance. The key idea is to sample lights by resampling instead of traversing light tree. To make a good balance between rendering quality and performance when resampling, we also propose a metric to measure lighting contribution. Experiments show that our approach generates comparable effects to the real-time stochastic lightcuts method and achieves significant performance improvement at the cost of modest rendering quality loss.