Inverse-Model Space for Feature Extraction from Absorption Spectra

Abstract

This report describes a methodology for spectrum-feature extraction from absorption spectra, obtained using reflectance measurements, including those obtained by diffuse reflectance (DR) and attenuated total reflectance (ATR) spectroscopies, based on reflectance theory and phenomenological function decomposition of reflectance. Specifically, this methodology entails spectrum feature-extraction using iterative spectrum adjustment by phenomenological backgrounds. Formulation of the feature-extraction methodology, based on inverse-analysis theory, is that of an inverse-model space, whose concept and properties are described. The inverse-model space, for spectrum feature-extraction, consists of numerical and analytical functions for non-unique iterative spectrum-background adjustment, which formally combines existing methods. In this study, the methodology defined by these combined methods, is extended to include truncated basis-function representations of dominant spectral features. In addition, results of spectrum-feature extractions demonstrating application of the inverse-analysis methodology are described.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 21, 2022
Accession Number
AD1167530

Entities

People

  • Andrew R. Shabaev
  • Samuel G. Lambrakos

Organizations

  • United States Naval Research Laboratory

Tags

Communities of Interest

  • Biomedical
  • Space

DTIC Thesaurus Topics

  • Absorption
  • Absorption Spectra
  • Caffeine
  • Case Studies
  • Chemistry
  • Classification
  • Coffee
  • Decomposition
  • Digital Signal Processing
  • Feature Extraction
  • Inverse Problems
  • Materials
  • Materials Science
  • Measurement
  • Military Research
  • Physics
  • Reflectance
  • Removal
  • Scattering
  • Signal Processing
  • Soil Science
  • Spectra
  • Spectroscopy
  • Standards

Readers

  • Calculus or Mathematical Analysis
  • Computational Modeling and Simulation
  • Spectroscopy.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Space