Path guiding method for wavefront path tracing: A memory efficient approach for GPU path tracers

Date

2024-3-04

Author

Yalçıner, Bora

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

252
views

25
downloads

Cite This

In this thesis, we propose a path guiding algorithm to be incorporated into the wavefront style of path tracers. As the wavefront technique of path tracers is primarily implemented on Graphics Processing Units (GPUs), the proposed method aims to leverage the capabilities of GPUs while reducing the hierarchical data structure usage and memory requirements necessary for path guiding methods. To achieve this, we propose a wavefront path guiding algorithm that only stores the outgoing irradiance (radiant exitance) on a single global Sparse Voxel Octree (SVO) data structure. Probability density functions required to guide the rays are generated on-the-fly using this data structure. The on-the-fly generation of the probability field is made practical by utilizing the ray-tracing hardware of the latest GPUs. Furthermore, the proposed approach significantly reduces the persistent memory requirements compared to other state-of-the-art path guiding techniques. Comparisons suggest that the proposed method requires less memory while maintaining similar results compared to the state-of-the-art techniques.

Subject Keywords

path tracing, GPGPU, ray tracing, light transport, hardware acceleration

URI

https://hdl.handle.net/11511/109165

Collections

Graduate School of Natural and Applied Sciences, Thesis