Comparison of the Accuracy of Rayleigh Rice Polarization Factors to Factors to Improve Microfacet Brdf Models
Abstract
Microfacet BRDF models assume that a surface has many small microfacets making up the roughness of the surface. Despite their computational simplicity in applications in remote sensing and scene generation, microfacet models lack the physical accuracy of wave optics models. In a previous work, Butler proposed to replace the Fresnel reflectance term of microfacet models with the Rayleigh-Rice polarization factor, Q, to create a more accurate model. This work examines the novel model that combines microfacet and wave optics terms for its accuracy in the pp and ss polarized cases individually. The model is fitted to the polarized data in each case, using the polarization factor Q, and the resulting fitted parameters are used to investigate whether parameters obtained using the ss and pp polarization may be used on the oppositely polarized or unpolarized data. Parameters fitted from pp polarization data t cross-term data better than those from ss polarization data for at least nine of the fourteen materials, indicating that more research must be done to make Q in the ss case more accurate. Model trends are determined to guide future work in refining polarimetric models.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2020
- Accession Number
- AD1104202
Entities
People
- Rachel L. Wolfgang
Organizations
- Air Force Institute of Technology