Radiance Filtering for Interactive Path Tracing
Those looking for details:
Unfortunately, the "radiance filtering" paper is still in review (after
almost over a year!), but the technique is described in my dissertation (Chapter 4), which can be downloaded from here (90MB). (Low quality version available here (5MB).)
You can also download some videos (800MB) that show the method in action and the supplemental material (850MB) with some additional stuff.
We present a method for noise reduction that is especially tailored to interactive progressive path tracing. The idea is to exploit spatial coherence in the image and reuse information of neighboring pixels. However, in contrast to image filtering techniques, we do not simply filter pixel values. Instead, we only reuse the incident indirect radiance of neighboring pixels in a radiance estimation step similar to stochastic progressive photon mapping. This novel approach significantly reduces the variance in indirect illumination without blurring details in geometry or texture. In equal time comparisons we often achieve higher quality than previous approaches. The primary use case of our algorithm is to provide fast, reliable previews of global illumination. It is also consistent, retains the conceptual simplicity of PT, is orthogonal to importance and stratified sampling, and is easy to integrate into existing renderers.
last modified by kschwenk on 02/09/14 09:31:47