photon_mapping

minimal but extensible header only implementation of photon mapping in C++.

Features

• Direct illumination with explicit light sampling
• Indirect illumination with final gathering
• Caustics photon map
• Reference path tracing integrator
• Load obj model

Requirements

• C++ (20>=)
• CMake (3.20>=)
• OpenMP
• spdlog
• Embree (>=3)

Build

CMake option	Description
BUILD_TESTS	build tests

git submodule update --init
mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
make

Run

place obj model under build and run

./examples/main

Structure

Name	Description
include/camera.h	ray generation from camera
include/core.h	basic data types
include/image.h	image
include/integrator.h	implementation of photon mapping, path tracing(for reference)
include/light.h	light
include/material.h	implementation of BRDF, BTDFs
include/photon_map.h	implementation of photon map with kdtree
include/primitive.h	primitive
include/sampler.h	random number generation, sampling utilities
include/scene.h	ray-scene intersection, model loading
include/triangle.h	point sampling on a triangle

Gallery

Cornell box without final gathering

Parameter	Value
number of photons	1000000
number of nearest neighbors	100
number of samples	100
final gathering	false

Cornell box with final gathering

Parameter	Value
number of photons	1000000
number of nearest neighbors	100
number of samples	100
final gathering	true

Cornell box with mirror spheres

Parameter	Value
number of photons	1000000
number of nearest neighbors	100
number of photons for caustics photon map	1000000
number of samples	100

Cornell box with water

Parameter	Value
number of photons	100000
number of nearest neighbors	100
number of photons for caustics photon map	10000000
number of samples	256

This model is available under models/

Externals

• spdlog
• Embree
• tinyobjloader

References

• Jensen, Henrik Wann. Realistic image synthesis using photon mapping. AK Peters/crc Press, 2001.
• https://pbr-book.org/3ed-2018/Light_Transport_III_Bidirectional_Methods/Stochastic_Progressive_Photon_Mapping#
• http://www.cs.cmu.edu/afs/cs/academic/class/15462-s12/www/lec_slides/lec18.pdf
• memoRANDOM
• McGuire Computer Graphics Archive
• Rendering Resources | Benedikt Bitterli's Portfolio
• Jensen, Henrik Wann. "Global illumination using photon maps." Eurographics workshop on Rendering techniques. Springer, Vienna, 1996.
• Veach, Eric. Robust Monte Carlo methods for light transport simulation. Stanford University, 1998.
• Christensen, Per H. "Faster photon map global illumination." Journal of graphics tools 4.3 (1999): 1-10.
• Hachisuka, Toshiya, Jacopo Pantaleoni, and Henrik Wann Jensen. "A path space extension for robust light transport simulation." ACM Transactions on Graphics (TOG) 31.6 (2012): 1-10.