Change Detection Analysis With Spectral Thermal Imagery.
Abstract
Spectral imagery offers additional information about a scene that can enhance an analyst's ability to conduct change detection. Automation of change detection is required to sift through countless images to identify scenes that have significant intelligence value. Change detection in spectral thermal imagery enables exploitation at night by taking advantage of the emissive characteristics of materials. Data collected from the Spatially Enhanced Broadband Array Spectrograph System (SEBASS) were used to investigate the feasibility of spectral thermal change detection in the long wave infrared (LWIR) region. This study used analysis techniques of differencing, histograms, and principal components analysis to detect spectral changes and investigate the utility of spectral change detection. Many artifacts can influence the sensitivity of change detection methods. Temperature dependence and gross registration errors greatly affect an analysts ability to make use of spectral thermal data for change detection; however, with effort, spectral changes were still detected with these data and suggest that the techniques would be useful once the undesirable characteristics are minimized.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1998
- Accession Number
- ADA356044
Entities
People
- Richard J. Behrens
Organizations
- Naval Postgraduate School