Non-parametric microfacet factor models for isotropic BRDF

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for representing the bidirectional reflectance distribution function (BRDF) of isotropic materials are cumbersome and inefficient, particularly when reflectance varies spatially, as they often rely on cumbersome 4D representations and assume simple analytic models for microfacet factors.

Innovation Solution

A non-parametric densely tabulated one-dimensional representation for isotropic materials is developed using an alternating weighted least squares (AWLS) method, which samples each factor function at discrete points without assuming a specific form, and applies a compressive weighting scheme to reduce the importance of bright highlights relative to darker areas.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a 4D representation is used for BRDF, then the reflectance can be accurately represented, but the representation becomes cumbersome and complex

Engineering Contradiction:
Improvereflectance representation accuracyVSAvoidBRDF representation complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the 4D BRDF representation into multiple 1D factor functions (normal distribution function, geometric factor, Fresnel factor) that can be independently tabulated and processed. This segmentation reduces the complexity of handling the full 4D data while preserving the essential reflectance characteristics through the microfacet model framework.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the 4D BRDF representation into a set of 1D factor functions by exploiting the microfacet model structure. This dimensionality reduction converts a cumbersome 4D tensor into multiple manageable 1D lookup tables, significantly reducing storage requirements and computational complexity while maintaining accuracy.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Ease of manufacture

If simple analytic models are assumed for microfacet factors, then the computation is simplified, but the accuracy of reflectance characterization is reduced

Engineering Contradiction:
Improvecomputation simplicityVSAvoidreflectance characterization accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent changes the approach from using fixed analytic models with predetermined parameters to using empirically measured 1D factor functions derived from actual material data. This allows the model to adapt to the specific characteristics of each material while maintaining the computational efficiency of the microfacet framework.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent creates empirical copies of the idealized analytic factor functions by measuring actual material reflectance and extracting 1D factor functions from the data. These empirical copies capture the true material behavior more accurately than theoretical models while preserving the computational benefits of the microfacet approach.

Inventive Principle:
Principle #26Copying

3Ease of manufacture

If equal weighting is applied to all data points, then the fitting is straightforward, but bright highlights dominate the error metric and reduce accuracy in darker areas

Engineering Contradiction:
Improvefitting process simplicityVSAvoidreflectance fitting accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent applies local quality by using weighted least squares fitting where the weight of each data point is inversely proportional to its reflectance value. This causes darker regions to have higher weights and brighter regions to have lower weights, allowing the fitting process to prioritize accuracy in darker areas while still capturing the overall reflectance characteristics.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the error metric by introducing reflectance-dependent weights in the least squares fitting process. This parameter change transforms the uniform weighting approach into a selective weighting scheme that emphasizes darker regions, thereby improving the overall accuracy of the reflectance characterization across different brightness levels.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3161797B1Non-parametric microfacet factor models for isotropic bidirectional reflectance distribution functions
Publication Date: 2020.01.08 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP3161797B1 patent drawingFigure 1
  • EP3161797B1 patent drawingFigure 2A~2B
  • EP3161797B1 patent drawingFigure 3

AI summary

A plurality of measured sample data points associated with reflectance on a surface of a material is obtained. A non-parametric densely tabulated one-dimensional representation for a plurality of factors in a microfacet model is generated, using the obtained sample data points.