Share Email Print

Proceedings Paper

Performance evaluation of linear-feature-mapping detection/classification algorithms using large SAR clutter dataset
Author(s): Lynne A. Tablewski; An Mei Chen; Clark R. Hendrickson; Xiaoli Yu; David L. Buck
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Using a linear feature mapping framework, a detector and classifier were developed for detecting and recognizing 2D target patterns in data collected by the SRI ultra wideband synthetic aperture radar (SAR), where limited prior knowledge of the target pattern and clutter statistics was known. Preliminary results on 0.1 square kilometers of the illustrative example site show that this linear feature mapping detector (LFMD) improves the receiver operating characteristic (ROC) curve by 1.5 orders of magnitude over the amplitude only 2-parameter constant false alarm rate (CFAR) detector, and the linear feature mapping classifier (LFMC) was able to classify 100% of the targets with no false alarms. In this report, the performances of both LFMD and LFMC are evaluated using ROC curves for 54 square kilometers of UHF SAR clutter data. Extensive experimental performance analysis using the large clutter data set shows that the LFMD and LFMC are very effective in detecting and classifying targets embedded in intense clutter background in UHF SAR data. Performance evaluation on 54 square kilometers of clutter data shows that the LFMD has only 0.3 false alarm per square kilometer and approximately 3 orders of magnitude improvement over the amplitude-only 2- parameter CFAR detector at probability of detection Pd equals 100%. The LFMC yields 100% correct classification by perfectly discriminating against the detected potential targets from the clutter data as nontargets.

Paper Details

Date Published: 5 June 1995
PDF: 10 pages
Proc. SPIE 2487, Algorithms for Synthetic Aperture Radar Imagery II, (5 June 1995); doi: 10.1117/12.210859
Show Author Affiliations
Lynne A. Tablewski, Naval Command, Control and Ocean Surveillance Ctr. (United States)
An Mei Chen, Science Applications International Corp. (United States)
Clark R. Hendrickson, Naval Command, Control and Ocean Surveillance Ctr. (United States)
Xiaoli Yu, Science Applications International Corp. (United States)
David L. Buck, Naval Command, Control and Ocean Surveillance Ctr. (United States)

Published in SPIE Proceedings Vol. 2487:
Algorithms for Synthetic Aperture Radar Imagery II
Dominick A. Giglio, Editor(s)

© SPIE. Terms of Use
Back to Top