Study and Investigation of Infrared Multispectral Target Cueing Techniques.
Abstract
This report presents the results of a study to analyze and evaluate multispectral infrared target cueing algorithms using an infrared data base (in the 1 - 14 micrometers region) of tactical targets against real backgrounds. Using a modified M7 line scanner, simultaneous multispectral imagery of five bands in the 1 - 14 micrometers region were collected. A description of the collected infrared data base is given including discussions of the motivating factors in determining the spectral regions employed and target and background optical properties. Also, M7 system performance characteristics are examined. Finally, a preliminary evaluation of two different approaches to multispectral target cueing using the collected data base was accomplished. The results of both approaches, the first involving classical maximum likelihood classification using in-scene training and the second involving unsupervised spectral/spatial clustering followed by discrimination based upon size are reported.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1979
- Accession Number
- ADA074799
Entities
People
- Jerry L. Beard
- Stephen R. Stewart
Organizations
- Environmental Research Institute of Michigan