25. accurate appearance preserving prefiltering for rendering displacement-mapped surfaces

Department: Computer Science & Engineering
Research Institute Affiliation: Center for Visual Computing
Faculty Advisor(s): Ravi Ramamoorthi

Primary Student
Name: Lifan Wu
Email: liw086@ucsd.edu
Phone: 858-531-9383
Grad Year: 2020

Abstract
Accurate multi-resolution representation for rendering complex surfaces remains a challenging problem. Micro-scale geometry induced by displacement maps produces variate illumination effects such as normal perturbation, shadowing and masking. Naive linear prefiltering of surfaces represented by high-resolution displacement maps weakens the object?s shadowing-masking effects and produces inaccurate appearance. We would like to propose a method to accurately prefilter displacement maps by jointly analyzing the visibility, the normal distribution function, and the BRDF. Our work leverages previous work on normal map filtering and microfacet theory. We optimize the surface scattering parameters to best match the ground truth appearance. We show our model results in surface rendering without aliasing.

Industry Application Area(s)
Computer Graphics

« Back to Posters or Search Results