CUDA Path Tracer

In this project, I implemented a CUDA-based path tracer capable of rendering globally-illuminated images very quickly. Path tracing is a computer graphics Monte Carlo method of rendering images of three-dimensional scenes such that the global illumination is faithful to reality. Basically, for every pixel of the image, we shoot a ray from the camera, find an intersection, and return a color based on the material, position of the objects & light source and so on after a few depths. As this process for every pixel is relatively similar, we can use GPU to run the programs simultaneously to make the path tracer faster.

Features

Ideal Diffuse surfaces
Perfectly specular-reflective surfaces
Stream Compaction
- By using stream compaction, we can remove the paths that have terminated;
Material sort
- Sort the rays/path segments so that rays/paths interacting with the same material are contiguous in memory before shading;
First bounce intersection cache
- Cache the first bounce intersections for re-use across all subsequent iterations;
Refraction with Frensel effects
- Refraction (e.g. glass/water) with Frensel effects using Schlick's approximation or more accurate methods;
Physically-based depth-of-field
- Physically-based depth-of-field by jittering rays within an aperture;
Stochastic Sampled Antialiasing
- Make the image smoother by add rays for sub-pixel sampling;
Motion blur
- Defining object motion, and motion blur by averaging samples at different times of iterations;

To see more, please check https://github.com/githublsx/Project3-CUDA-Path-Tracer.