Remote Material Identification Using Multi-Angle Spectral Signatures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote material identification methods using spectral signatures often yield erroneous results when dealing with non-Lambertian materials, as they treat all targets as Lambertian, failing to account for the anisotropic reflectance and emissivity of non-Lambertian surfaces.
Innovation Solution
The method involves determining spectral signatures from multiple angles and comparing them to predicted signatures that account for the known anisotropy of non-Lambertian candidate materials, using bi-directional reflectance distribution functions and analyzing uncertainties in multi-angle signatures to enhance the confidence in material identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If all targets are treated as Lambertian in remote material identification, then the identification process is simple, but the identification accuracy deteriorates for non-Lambertian materials
Solution Approach 1:
The patent changes the parameter assumption from Lambertian (isotropic) reflectance to non-Lambertian (anisotropic) reflectance by incorporating bidirectional reflectance distribution functions (BRDF). This allows the identification system to account for directional variations in reflectance and emissivity, thereby improving identification accuracy for non-Lambertian materials while maintaining a systematic approach to the increased complexity
Solution Approach 2:
The patent introduces angular dimensionality by acquiring spectral signatures from multiple viewing angles. Instead of single-angle measurement, the system collects multi-angle data and compares it against multi-angle predicted signatures, adding the angular dimension to the identification process to capture anisotropic properties of non-Lambertian materials
2Measurement precision
If multi-angle spectral signatures are used for non-Lambertian materials, then identification accuracy improves, but the complexity of the identification process increases
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing predicted spectral signatures for candidate materials at multiple angles before actual identification occurs. The BRDF-based predicted signatures are prepared in advance, allowing the identification system to simply compare measured multi-angle signatures against these pre-computed references, thereby reducing the computational complexity during the actual identification process
Solution Approach 2:
The patent introduces bidirectional reflectance distribution functions (BRDF) as an intermediary that bridges the gap between material properties and observed spectral signatures. The BRDF model serves as a mediator to predict how candidate materials would appear from different angles, enabling the system to handle multi-angle data without directly solving complex inverse problems for each measurement
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach increases the probability of correct identification and decreases false matches by accurately accounting for the anisotropic properties of non-Lambertian materials, providing more reliable remote material identification.
Implementation Method 1
Different materials absorb, reflect and emit differently, depending on wavelength. The spectral radiance of a target... can be measured using various types of sensors
Implementation Method 2
A non-Lambertian surface does not scatter incident electromagnetic radiation equally in all directions. Examples of non-Lambertian surfaces include those that are backscattering, meaning that the light scatters predominantly toward the illumination source
Data Source
AI summary
In one example of a method for remote identifying a non-Lambertian target material, a spectral signature for a target is determined from each of at least two different sets of imagery acquired at different angles, and compared to a predicted signature for a candidate material for each of the at least two different angles. The predicted signatures take into account the known anisotropy of reflectance, and thus also radiance, of the candidate material.