Multi-Sensor Feature Level Fusion
Abstract
The Multi-Sensor Feature Level Fusion (MSFLF) program is intended to improve automatic target recognition (ATR) performance by the use of multi- sensor feature techniques. MSFLF algorithms are evaluated using multi-sensor (8- 12m FLIR, 94GHz MMW radar and 10.6 micron CO2 laser radar) data and imagery. This report gives the results of target detection and identification using the MSFLF training IR imagery. A set of image quality measures are defined and computed over the imagery. These quality measures suggest that performance should be better for the training imagery than for the test imagery. An overview of the FLIR and MMW radar algorithms, detailed descriptions of the IR target detection and identification algorithms are presented in a separate proprietary report.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 22, 1991
- Accession Number
- ADA237106