Evaluation of the Statistics of Target Spectra in Hyperspectral Imagery (HSI)

Abstract

The majority of spectral imagery classifiers make a decision based on information from a particular spectrum, often the mean, which best represents the spectral signature of a particular target. It is known, however, that the spectral signature of a target can vary significantly due to differences in illumination conditions, target shape, and target material composition. Furthermore, many targets of interest are inherently mixed, as is the case with camouflaged military vehicles, leading to even greater variability. In this thesis, a detailed statistical analysis is performed on HYDICE imagery of Davis Monthan Air Force Base. Several hundred pixels are identified as belonging to one of eight target classes and the distribution of spectral radiance within each group is studied. It has been found that simple normal statistics do not adequately model either the total radiance or the single band spectral radiance distributions, both of which can have highly skewed histograms even when the spectral radiance is high. Goodness of fit tests are performed for maximum likelihood normal, lognormal, gamma, and Weibull distributions. It was discovered that lognormal statistics can model the total radiance and many single-band distributions reasonable well, possibly indicative of multiplicative noise features in remotely sensed spectral imagery.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2000
Accession Number
ADA384317

Entities

People

  • Joel C. Robertson

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Air Platforms
  • Sensors
  • Space

DTIC Thesaurus Topics

  • Air Force
  • Air Force Facilities
  • Aircrafts
  • Computer Programs
  • Data Science
  • Detection
  • Detectors
  • Distribution Functions
  • Goodness Of Fit Tests
  • Hyperspectral Imagery
  • Information Science
  • Materials
  • Random Variables
  • Spectra
  • Statistical Analysis
  • Statistical Distributions
  • Statistics

Readers

  • Image Processing and Computer Vision.
  • Regression Analysis.
  • Statistical inference.