Institute of Computer Graphics, Vienna University of Technology, 2000. - 123 pp.
Достаточно подробно описаны методы Монте-Карло для глобального освещения (в том числе и метод русской рулетки), приведены описание задачи глобального освещения, описание моделей освещения, реализация стохастической трассировки лучей на языке C++ с диаграммой классов и диаграммой последовательностей.Introduction.
Global pass.
Local pass.
Tone mapping.
Global illumination problem.
The rendering equation.
Measuring the radiance.
The potential equation.
Measuring the potential.
The rendering problem.
Geometry of the surfaces.
Bi-directional Reflection Distribution Functions.
Lightsources.
Measuring devices.
Numerical solution of the rendering equation.
Error measures for numeric techniques.
Properties of the rendering equation.
Classification of the solution techniques.
Optical material models.
Diffuse reflection.
deal, mirror-like reflection.
deal refraction.
Non-ideal, specular reflection.
Phong reflection model and its modifications.
Cook-Torrance model.
Solution strategies for the global illumination problem.
nversion.
Expansion.
Expansion of the rendering equation: gathering walks.
Expansion of the potential equation: shooting walks.
Merits and disadvantages of expansion methods.
teration.
Analysis of the iteration.
Analytical solution of the rendering equation.
Scenes with constant radiance.
Scenes with constant reflected radiance.
Finite-element methods for the Global Illumination Problem.
Galerkin’s method.
Point collocation method.
Finite element methods for the diffuse global illumination problem.
Geometric methods for form factor computation.
Numerical quadrature for high dimensional integrals.
Monte-Carlo quadrature.
Quasi-Monte Carlo quadrature.
Error Analysis for integrands of finite variation: Koksma-Hlawka Inequality.
Generation of the sample points.
Generation of low-discrepancy sequences.
mportance sampling.
Generation of a random variable with a prescribed probability density.
mportance sampling in quasi-Monte Carlo integration.
Metropolis sampling.
Application of the VEGAS algorithm.
Random walk solution of the global illumination problem.
Why should we use Monte-Carlo expansion methods?
.
Quasi-Monte Carlo quadrature for the rendering equation.
ntegrating functions of unbounded variation.
mportance sampling for the rendering equation.
BRDF sampling.
Lightsource sampling.
Sampling the lightsources in gathering random walks.
mportance sampling in colored scenes.
Multiple importance sampling.
Handling infinite-dimensional integrals.
Russian roulette.
Russian roulette in quasi-Monte Carlo quadrature.
Review of random walk algorithms.
Gathering-type random walk algorithms.
Ray-casting.
sibility ray-tracing.
Distributed ray-tracing.
Path-tracing.
Shooting-type walks methods.
Photon tracing.
Light-tracing.
Random walks for the radiosity setting.
Bi-directional random walk algorithms.
Bi-directional path-tracing.
Metropolis light transport.
Photon-map.
nstant radiosity.
Global methods.
Multi-path method using global random lines.
Global ray-bundle tracing.
Preprocessing the point lightsources.
teration solution of the global illumination problem.
Why should we use Monte-Carlo iteration methods?
Formal definition of stochastic iteration.
Other averaging techniques.
Can we use quasi-Monte Carlo techniques in iteration?
Review of stochastic iteration algorithms.
Stochastic iteration for the diffuse radiosity.
Stochastic radiosity.
Transillumination radiosity.
Stochastic ray-radiosity.
Definition of the random transport operator for the non-diffuse finite-element case.
Single ray based transport operator.
Stochastic iteration using ray-bundles.
mplementation of the path-tracing algorithm.
ector module.
Point3D class.
Transformation class.
Container module.
Color module.
Material models.
Diffuse material class.
deal mirror class.
deal refracting material class.
Specular material class.
General material class.
Light module.
Emitter class.
Positional light class.
Model module.
Primitive class.
Object class.
rtual world class.
Camera module.
Scene module.
Scene class.
Dynamic model of path tracing.
Finding the visible primitive.
Detecting the visible lightsources.
Direct lightsource computation.
Pathtracing.
Rendering complete images.
