Wondering how hair is stored and formatted in the memory. I understand that a segment of hair can be approximated as a cylinder. How about an entire hair? Is it represented as combination of many cylinders? And how about all hair? Also, given the special structure of hair, how to build hierarchical data structure to store it to accelerate intersection computation?
The main difference between Kaijiya-Kay Model and other hair-shading model is that it uses hair strand tangent instead of normal in lighting equations.