Measurement of the Optical Properties of Complex Surfaces James - PowerPoint PPT Presentation

The Prediction and Measurement of the Optical Properties of Complex Surfaces James Jafolla Surface Optics Corporation ITBMS 2018

Outline • Introduction • BRDF Definition • BRDF Models – Micro-Optic Surface Model – Sandford-Robertson Model • BRDF Measurements • BRDF Model-Measurement Comparison • Implications for Signature Analysis

Introduction • Development of advanced CGI rendering tools requires better understanding of light scattering from surfaces • Virtual prototyping for signature analysis and system design requires accurate radiometric analysis • Most current computer analysis relies on simplified assumptions of surface optical properties • Phenomenological ray tracing analysis provides accurate prediction of optical properties

Bidirectional Reflectance Distribution Function (BRDF) BRDF: ( )      , , i r ( )   ( ) dN ,  =       r r , , cos d ( )  i r i i N i i ( )   DHR: D i ( ) N r  =          ( ) , , cos sin d d r r r r i  N i i ( ) =  D  i  i (These apply to isotropic surfaces; also,  r = Incident Zenith Angle    −  here.) r i  = Reflected Zenith Angle    =  = Reflected Azimuth Angle For Lambertian Diffuse Surface D

Modeling the BRDF of Complex Surfaces • Many surfaces exhibit a complex surface features • Large scale features exhibit multiple reflections and blocking and shadowing • Small scale features based on homogeneous material optical properties, surface roughness • Non-homogeneous mixtures of materials

Micro-Optic BRDF Calculation • Approach defines Micro-Optic facet model of surface • Ray-tracing includes reflections and blocking between facets • Sandford-Robertson model for facet optical properties

Geometrical Surface Structures • Structured Surfaces for Thermal Design • Ref. Seigel R. and Howell J., Thermal Radiation Heat Transfer , McGraw-Hill, 817pp, 1972 • Arbitrary Structure Surfaces Represented as Faceted Wire-Frame Models • Arbitrary Surface Coatings on Facets

Pyramid Surface Definition Machined Surface

SOC-210 Bi-Directional Reflectometer Unpolarized BRDF Bidirectional Transmittance Distribution Function (BTDF) Automated Full Hemispherical Mapping:  i = 0  through +85   r = 0  through +85   i = 0  - 350   r = 0  - 345  Spectral Coverage: Sources 0.35 – 1.6  m quartz halogen 1.5 – 16.0  m SiC glower source Assorted bandpass filters Detectors 0.35 -1.0  m Silicon detector 1.0 – 1.6  m InGaAs detector 2.0 – 14  m dual InSb/MCT

BRDF Measurements

Sandford-Robertson Model • Based on the separation of the spectral and directional dependence of the total BRDF       l =      l ( , ; , ; ) f ( , ; , ) ( ) i i r r r i i r r     =     +     f ( , ; , ) f ( , ; , ) f ( , ; , ) r i i r r S i i r r D i i r r • Fits four parameters to the BRDF  D ( l ) = Diffuse Spectral Reflectance e ( l ) = Spectral Emissivity b = Grazing Angle Reflectivity e = Width of Specular Lobe

Sandford-Robertson Model Fit to Measurements

Analytical SR BRDF Models Good Specular Mid Specular Bad Specular

Model/Measurement Comparison Theta Incident 40 deg Prediction Measurement

Study of Grooved Surface • 30 degree right triangle surface wedge • Movie generated of BRDF for high resolution steps in Phi Incident for various Theta Incident angles • Good specular surface BRDF

Incident Angle 35 Degrees

Incident Angle 55 Degrees

Incident Angle 65 Degrees

Implications For Signature Analysis • Regular (structured) surface features produce multiple BRDF lobes • Lobes move predictably as a function of all four incident and scattered angles • Inclusion in signature models using high resolution data tables and interpolation possible • Parameterized models have been attempted

Hilgers BRDF Model • Developed to Model Materials with Multi-Lobe BRDF Features • Gaussian Lobe Shape • Closed Form Solution - No Iterative Fitting • Lobes “ Tracked ” to Define Significant Features – i.e., Lobe Bifurcation or Coalescence • Rapid Generation of BRDF for Rendering Applications ˆ ˆ  −   = ( k , k ) Ae i r  =  −  +  −   −  +  −  2 2 a ( ) 2 b ( )( ) c ( ) r r r r r r r r A determined from BRDF peak  , r  determined from first moment of BRDF r a , b , c determined from second moment of BRDF

Multi-Lobe BRDF Comparison Q i =10  i =0 Gaussian Model MicroOptic BRDF Q i =50  i =40

Full BRDF Provides Realistic Rendering Courtesy: George Borshukov, ESC Inc.

Computer rendered character from a “famous” movie. photo rendered Courtesy: George Borshukov, ESC Inc.

Conclusions • Simple parameterized BRDF models not well suited for complex surface treatments • Micro-Optic BRDF model provides ability to represent complex, non-homogeneous surfaces – Geometric facet model of surface – Multiple bounce ray-tracing calculation – BRDF of homogeneous facets computed from Sandford- Robertson model • Good agreement of BRDF calculations to measurements for a manufactured structured surface • More work needed for including these effects into signature calculations