Progressive Photon Mapping 최인창(Inchang Choi) KAIST (Korea Advanced Institute of Science and Technology) 1

Intro: MC based methods can simulate specular and diffuse material Problems: Caustics (LS+DE) SDS paths (LSDSE) Can simulate both specular and diffuse material Problem1: caustic (definition) ->example Problme2: SDS path -> example (swimming pool) (sampling problem)

Intro: Photon Mapping Two pass algorithm Problem: First pass: Photon Tracing Second pass: Ray Tracing using the photon maps Problem: Scenes dominated by caustics ->We don’t have infinite memory Ray tracing using photon mapps First pass: photon tracing to make two photon maps Second pass: ray tracing using photon map Problem: if the scence is dominated by causitics illumination (caustics usually are simulated by directily visualizing photon maps, and it requires too much memory) -> accuracy is regulated by memory available.

Intro: Progressive Photon Mapping An extension of Photon Mapping Include SDS paths with arbitrary accuracy without infinite memory Multi-pass algorithm First Pass: Ray Tracing Subsequent Passes: Photon Tracing A progressive photon mapping algorithm simulated global illumination including SDS paths with arbirtray accuracy without infinite memory. Multipass alogrithm – 1. ray tracing 2. photon tracing

Photon Mapping

Photon Mapping: First Pass: Photon Tracing Shoot photons from the light source -using MC technique and Russian Roulette -store photons in K-d tree Position, incident direction, flux Manage two photon maps Caustic map Global photon map Caustic map photons Global map photons From John C. Hart’s ppt

Photon Mapping: Second Pass: Ray Tracing Trace rays Calculate illumination using two photon maps DA = pr2 Use the nearest N photons to estimate L at point x From John C. Hart’s ppt

Photon Mapping: Second Pass: Ray Tracing Direct Illumination Diffuse Indirect Illumination Specular Illumination Caustic Illumination From Michael Kaiser and Christian Finger’s ppt

How can it be improved? 1. Make the number of photons very large (lim n->∞) Use an Iterative method 2. Reduce the radius r (limt r-> 0) Reduce the radius every iteration

Progressive Photon Mapping

Progressive Photon Mapping First Pass: Ray Tracing Shoot a ray from eye to the scene Trace the path until the first non-specular surface Store the hit point if the surface has non- specular(diffuse) component in the BRDF

Progressive Photon Mapping First Pass: Ray Tracing A Hit Point Structure These are updated each subsequent passes

Progressive Photon Mapping Subsequent Pass: Photon Tracing In each Pass Trace a given number of photons into the scene

Progressive Photon Mapping Subsequent Pass: Photon Tracing After each Pass Loop all hit points :Find the photons within the radius :refine the estimate of the illumination (Reduce R, increase N, update τ)

Progressive Photon Mapping Radius Reduction

Progressive Photon Mapping Flux Correction

Progressive Photon Mapping Radiance Evaluation

A Video Result http://www.youtube.com/watch?v=zg9NcCw53iA

Result Comparison

Thank you for listening Any Questions?