cs184-final-project

Physically-Based Fire Rendering

We plan to implement physically-based fire rendering using detailed simulations of flame dynamics. Our proposed path tracer will account for black body radiation, spectral properties, refraction, and visual adaptation (among other factors).

The three team members are Justin Chen, Balaji Veeramani, and Michael Wang.

Problem Description

The problem we are trying to solve is the ability to realistically render fire. The ability to realistic render flames is important because fire appear in many artistic rendering applications (for example, games and movies). The problem is challenging because flames arise from complex electromagnetic waves interact.

We intend to solve the problem by implementing a physically-based rendering algorithm. The algorithm we are implementing is contrasted with ray-tracing and raster-based techniques, which fail to achieve the level of accuracy we desire. The paper describing the algorithm is listed in the resources section below.

Goals and Deliverables

To create realistic fire simulation, we divide the project into the following features:

We think we can implement these features because we can implement the features by extending our existing path tracer.

Below is a photos illustrating our expected result.

We will measure the realism of our pathtracer by comparing rendered images to real-life photographs, as well as renders from the original paper. We will not measure the performance of the pathtracer because real-time rendering is a non-goal of the project.

We plan to deliver a pathtracer that can render (somewhat) realistic flames. The semi-realistic flames can be constructed with either a physically-based algorithm or a non-physically-based shader. If everything goes well, then we hope to implement all parts of a physically-based flame rendering algorithm.

Schedule

Below are the tasks we hope to complete over the next four weeks:

Resources

Here are various resources we plan to utilize: