BROWSE
Detail
Conference Paper
Thread-based BRDF Rendering on GPU
Cited - time in 
Share 
Share
- Authors
- Soon Hyun Kim, Min-Ho Kyung, Joo-Haeng Lee
- Issue Date
- 2010-09
- Citation
- Pacific Conference on Computer Graphics and Applications (Pacific Graphics) 2010, pp.1-9
- Language
- English
- Type
- Conference Paper
- Abstract
- Rendering BRDF surfaces have been intensively studied to produce physically plausible appearance of surface materials. Illumination at a surface point is formulated as an integral of a BRDF producted with incident radiance over the hemi-sphere domain. One popular method to compute the integral is Monte Carlo integration which estimates it with a sum of integrand evaluated at stochastically sampled rays. Although its simple nature is practically attractive, it has a serious drawback of noise artifacts in 3D rendering. In this paper, we propose a novel noisefree Monte Carlo rendering algorithm running on a GPU in real-time. The main contribution is a new importance sampling scheme making a noise-free image without increasing the sample count. For each evenly spaced latitude angle of eye ray, denoted by θ, incident rays are sampled with a PDF derived from the target BRDF lobe. We use a force-based update method to make incident rays to be consistently sampled along consecutive θ's. Finally, corresponding incident rays are linearly connected to form a smooth curve, called a sample thread. In rendering, the ray samples for integration are obtained as points on the threads specified by the eye ray at a surface point. Since sample rays obtained from threads are locally consistent on a target surface, the image noise manifested by estimation variance is not found. A thread set is precomputed for a specific BRDF before rendering, so that no sampling overhead is imposed on the GPU. According to our experiments, at most 100 threads are necessary for most measured BRDFs to achieve a plausible quality, which enables real-time performance.
- KSP Keywords
- 3D Rendering, Force-based, GPU rendering, Image noise, Monte Carlo integration, Monte Carlo rendering, Noise Artifacts, Sampling scheme, Surface point, importance sampling, real-time performance