Hair Simulator - CS184
Our group aims to statically render realistic hair and (ideally) fur. The reflectance models are hard to simulate because hair and fur have complex light-scattering properties that make it difficult to emulate via ray-tracing. Hair/fur also have many physical properties like thickness, length, bend radius, that make it hard to render when applying physics(wind, motion). We are going to use the Marschner model for rendering hair and Lingqi's state of the art fur reflectance model as our primary resources.
For this project, we plan to deliver a series of images generated with the Mitsuba Renderer. We will first use the renderer to generate some basic images of a head of hair, without any advanced lighting models. Then after implementing the Marschner model, we will generate several images of much more realistic human hair to compare to the basic renderings.
When working on this project, we will have to figure out how to render realistic renderings of hair. Therefore, We have a couple stretch goals depending on the difficulty of the hair rendering and what issues we come across. First, we would like to use the Linq-Qi model for fur to try to generate realistic fun renderings for animals. In case this is also successful, we would also like to set up some sort of interactive real time rendering demo using WebGL or OpenGL.
For the fur rendering deliverables, we will try to generate several images of one or two animals. This process seems more difficult, so we may have multiple image sections showcasing our progress on the fur rendering. Then we would like to set up some sort of webpage using OpenGL to have a dynamic interactive rendering of our hair and/or fur models.
At this time, it is difficult to measure the quality of our renderings using any objective methods, since it is more about a realistic appearance. Since we are only using one model for each render, we also can’t check for improved performance, since we have no baseline. Therefore, the quality of our images will be mostly subjective.
Week 1: Familiarize ourselves with the Mitsuba render, find more models, resources, read through all of the research papers on hair rendering
Week 2: work on implementing the Marschner model for rendering human hair
Week 3: Finish Marschner model and move on to fur rendering.
Week 4: Finish fur rendering and (if there’s still more time) make a dynamic fur/hair renderer in OpenGL or WebGL
http://www.eugenedeon.com/wp-content/uploads/2014/04/egsrhair.pdf http://people.eecs.berkeley.edu/~lingqi/publications/paper_fur2.pdf http://nickshen.me/184proj/final/ http://mitsuba-renderer.org/
Ubuntu 18.04 Windows 10 MacOs Mojave
Mitsuba Renderer OpenGL (Maybe)