Image-based acquisition and modeling of polarimetric reflectance

Abstract

color albedos, captured in five wavelength ranges covering the visible spectrum. We demonstrate usage of our data-driven pBRDF model in a physically based renderer that accounts for polarized interreflection, and we investigate the relationship of polarization and material appearance, providing insights into the behavior of characteristic real-world pBRDFs. CCS Concepts: • Computing methodologies → Image and video acquisition; Reflectance modeling.

Merit functions and measurement schemes for single parameter depolarization models

“…All vectors point in the direction light travels (i.e. k-vector direction) [10,11]. The object surface normal is n which also referred to as a macronormal to distinguish from microfacet normals.…”

“…In computer graphics literature, the use of a halfway vector h is common in the implementation of microfacet BSDF models [8,10,11,22,35]. For backscattering events, the halfway vector bisects the incident and reflected ray [22].…”

“…Polarized BSDF models of rough surfaces have been explored as both forward models for material rendering [9][10][11][12] and inverse models for material recognition [13][14][15]. Prior work has shown how applying pBSDF microfacet models to a Fresnel reflection component can improve the realism of a rendered scene [10][11][12]. Microfacets are planar structures on a larger macrosurface that only reflect in the specular direction.…”

“…Microfacets are planar structures on a larger macrosurface that only reflect in the specular direction. Recent pBSDF models have experimented with removing the ideally depolarizing component and replacing it with a diffuse or partiallydepolarized component to improve measurement agreement [10,11]. A recent pBSDF model by Kondo et al experiments with adding a third diffuse component instead of replacing the ideally depolarizing component to further improve model and measurement agreement [12].…”

Evaluating the Utility of Mueller Matrix Imaging for Diffuse Material Classification

“…In remote sensing, classification of ground targets from radar polarimetry combined with statistical decision-making techniques are well established [14]. Material classification for visible and near visible Mueller matrix imaging has been tested with a small field of view bench top imaging systems with an emphasis on post-processing algorithms and polarized light scattering models [15][16][17][18]. As the diffuse reflectance of a material increases, the degree of polarization of the scattered light decreases; this is known as the Umov Effect [19].…”