Jon Sporring
Photon Differentials in Time and Space
Sampling the light space for global illumination is an algorithmic challenge due to the size of the domain, the number of objects typically contained in a scene as well as the number of physical effects to take into account. Photon mapping has been introduced in the literature as an algorithmic technique for estimating the distribution of light in a scene as seen from a camera, but a central problem in photon mapping (and other sampling techniques) is aliasing and noise. Smoothing is a central concept to avoid aliasing and reduce noise, and we present photon differentials as a technique for geometry and motion driven smoothing by modeling the relation between light sources, objects and cameras. This talk will give an introduction to photon mapping, and photon differentials in time and space.