Reflectance Spectrum Evaluation Using Iterative Envelope Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating hyperspectral data struggle to accurately determine characteristic absorption features, especially complex or composite features, due to limitations in prior knowledge-based expert systems and methods like convex hull and Alpha Shapes, which fail to capture important features like the 'green peak' and 'red edge', and are inadequate for rare earth minerals and fine absorption features.
Innovation Solution
A system that uses a combination of smoothing filters and iterative linear interpolation to determine local maxima and form upper envelopes, allowing for the identification of characteristic absorption features without relying on prior knowledge or databases, and can handle complex features and superimposed absorption bands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If prior knowledge-based expert systems and methods like convex hull and Alpha Shapes are used to evaluate hyperspectral data, then the evaluation process is simplified and can be performed with existing databases, but important characteristic features like the 'green peak' and 'red edge' are not captured and complex or composite absorption features are missed
Solution Approach 1:
The evaluation process is divided into multiple processing stages: initial smoothing with first smoothing parameters, identification of local maxima and minima, formation of upper and lower envelopes, subtraction to obtain residual spectrum, and iterative refinement with second smoothing parameters. This segmentation allows each stage to focus on specific features without being constrained by prior knowledge databases, thereby capturing complex and composite absorption features that single-stage methods miss.
Solution Approach 2:
The method employs iterative refinement where the residual spectrum obtained from the first evaluation cycle is subjected to further processing with adjusted smoothing parameters. This dynamic, multi-pass approach allows the system to progressively reveal absorption features at different scales and complexities, improving detection precision without requiring predetermined knowledge of what features to look for.
2Measurement precision
If manual specification of absorption bands is required for accurate material identification, then the precision of feature detection improves, but the time and complexity of the evaluation process increases significantly
Solution Approach 1:
The system performs self-service by automatically identifying characteristic absorption features through mathematical operations on the reflectance spectrum itself. The method uses automated envelope formation, local extremum detection, and iterative residual analysis to identify features without human intervention or reference to external databases. This self-service capability maintains high detection precision while eliminating the time-consuming manual specification process.
Solution Approach 2:
The method employs parameter changes by using different smoothing parameters in different processing stages. First smoothing parameters are applied initially, then after envelope subtraction, second smoothing parameters are applied to the residual spectrum. This dynamic adjustment of parameters allows the system to adapt to different spectral features automatically, maintaining precision without manual intervention.
3Loss of information
If hyperspectral sensors record a large number of wavelength channels to capture detailed spectral information, then the completeness of spectral data improves, but the difficulty of evaluating and interpreting the data increases since humans can only perceive three wavelength channels
Solution Approach 1:
The method extracts the essential characteristic absorption features from the large hyperspectral dataset by identifying local maxima and minima, forming envelopes, and analyzing residuals. This extraction process distills the information-rich but complex hyperspectral data into meaningful absorption features that can be directly used for material identification, effectively removing the evaluation complexity while preserving spectral information completeness.
Solution Approach 2:
Instead of trying to interpret the entire continuous spectrum directly (which is difficult for humans), the method inverts the approach by focusing on the deviations from the envelope - the absorption features. By subtracting the upper and lower envelopes and analyzing the residual spectrum, the method highlights the characteristic features while filtering out the continuous spectral information, making the data evaluation tractable while preserving all spectral information.
4Measurement precision
If databases or libraries with prior knowledge are used to complement human evaluation of reflectance spectra, then the evaluation accuracy improves by comparing with known spectra, but the system cannot identify materials or features not recorded in the databases
Solution Approach 1:
The method provides a universal evaluation approach that works for any reflectance spectrum regardless of whether it matches known materials in databases. By using mathematical operations (smoothing, envelope formation, residual analysis) that are applicable to any spectral shape, the system can identify characteristic features of known and unknown materials equally effectively. This universality enables both accurate evaluation of known materials and discovery of new materials not in databases.
Solution Approach 2:
The system performs self-service by deriving all necessary information directly from the input spectrum through automated mathematical processing. It does not rely on external databases or prior knowledge but instead uses the spectral data itself to identify characteristic absorption features. This self-service capability ensures high adaptability to unknown materials while maintaining evaluation accuracy through rigorous mathematical analysis.
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a system and a method for evaluating reflectance spectra and determining characteristic absorption features in these spectra. The system comprises an input unit, a computing unit, and an output unit, as well as a device for recording the reflectance spectra. The reflectance spectra are recorded by the device and forwarded to the computing unit for processing via the input unit, where they are evaluated. The method is carried out using the system according to the invention.