Abstract
A modified pBRDF model with a diffuse scattering component is applied to estimate the complex refractive index, slope variance roughness, and diffuse scattering coefficients of object surfaces from time sequences of polarimetric images. The approach is used for the first time to produce parameter-based images from multispectral Stokes imagery of outdoor target scenes collected by the Ground Multiangle Spectro-Polarimetric Imager. The images of the estimated surface parameters show distinctions between different objects in the scenes and the parameter values are consistent with reasonable expectations for the object surfaces.
Highlights
Optical polarimetry allows the measurement or estimation of surface material properties and supports remote sensing applications [1] such as material classification [2], shape extraction [3], and target recognition [4]
We summarize the modified polarimetric bidirectional reflectance distribution function model and the parameter estimation approach used for this work [14]
The images generated from the estimated surface parameters show distinctions between the same objects in the scenes as the DOP images
Summary
Optical polarimetry allows the measurement or estimation of surface material properties and supports remote sensing applications [1] such as material classification [2], shape extraction [3], and target recognition [4]. Following the approach of Hyde et al, Yang et al [12] jointly estimated the refractive index and surface roughness of several non-specular samples from a series of DOP observations Their model and estimation approach are only valid for highly reflective materials since optical absorption inside the material is neglected. In recent studies [13,14], we describe two independent approaches for estimating the refractive index, roughness and diffuse scattering coefficient of a surface from measurements of the degree of polarization (DOP) and the Stokes parameters, respectively. The parameter estimation results using both approaches are encouraging, and synthetic simulation results of DOP and Stokes parameters generated by the proposed models show close correspondence to measured values These proposed approaches significantly simplify the surface parameter estimation process without losing accuracy especially for remote sensing where target surfaces are typically rough compared to the incident wavelength. The polarimetric data for this work was collected with the Ground Multiangle SpectroPolarimetric Imager (Ground-MSPI) [15,16], which is a camera system developed to provide Stokes parameter images at narrowband wavelengths of 470 nm, 660 nm and 865 nm [17,18,19,20,21,22]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
