Bidirectional light transport with vertex merging

SIGGRAPH Asia 2011 (technical sketch)

Left: The SIGGRAPH Souvenirs scene reference image. Middle: Insets showing the quality achieved by 4 different algorithms in the same time (200 seconds): Combined vertex merging with bidirectional path tracing (VM+BDPT), vertex merging only (VM), progressive photon mapping (PPM), bidirectional path tracing (BDPT). The new VM+BDPT algorithm combines various path sampling techniques to produce the image with the lowest overall error. Right: A path segment generated by a vertex connection (top) and by vertex merging (bottom).

Abstract

We present vertex merging – a bidirectional path sampling technique for Monte Carlo light transport integration. Vertex merging is simple and more computationally efficient for specular-diffuse-specular effects than the currently available techniques in bidirectional path tracing. It brings the advantages of photon mapping to the path integral framework, while avoiding the concept of density estimation altogether. This makes it possible for the first time to quantitatively reason about the efficiency of two rendering approaches that have been historically considered conceptually different. The practical result is a combined bidirectional rendering algorithm that efficiently handles a wide variety of lighting conditions, ranging from direct illumination and diffuse inter-reflections to the notoriously problematic reflected caustics. This algorithm also has a higher order of convergence than progressive photon mapping.

Downloads and links

BibTeX reference

@inproceedings{Georgiev:2011:VM,
  author = {Georgiev, Iliyan and K\v{r}iv\'{a}nek, Jaroslav and Slusallek, Philipp},
  title = {Bidirectional light transport with vertex merging},
  booktitle = {SIGGRAPH Asia 2011 Sketches},
  series = {SA '11},
  year = {2011},
  isbn = {978-1-4503-1138-0},
  location = {Hong Kong, China},
  pages = {27:1--27:2},
  articleno = {27},
  numpages = {2},
  url = {http://doi.acm.org/10.1145/2077378.2077412},
  doi = {http://doi.acm.org/10.1145/2077378.2077412},
  acmid = {2077412},
  publisher = {ACM},
  address = {New York, NY, USA},
}