Feature Extraction Using Attributed Scattering Center Models for Model-Based Automatic Target Recognition (ATR)
Abstract
"Feature Extraction using Attributed Scattering Center Models for Model-Based Automatic Target Recognition." The primary research goal of the program was to develop fundamental understanding and advanced signal processing techniques for feature extraction to support feature-based automatic target recognition (ATR) systems employing synthetic aperture radar. This report summarizes the major technical accomplishments that were realized. We developed a set of attributed scattering center models for SAR ATR whose model primitives that balance between modeling fidelity and estimation accuracy. We developed computationally-efficient algorithms for automatic feature extraction of attributed scattering center features from complex SAR image-domain data. We analyzed feature uncertainty and derived analytical uncertainty bounds. We implemented stand-alone match scoring methods to evaluate target discriminability and feature estimation tradeoffs. We developed STAP/SFAP-Based Adaptive Antennas. We developed techniques for understanding rough surface scattering. We developed ultrawide bandwidth antennas, and slot array antennas with wide scan angles. Finally, we increased the U.S. technology base by training of graduate students and by disseminating research through technical publications and presentations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2005
- Accession Number
- ADA444563
Entities
People
- Inder J. Gupta
- Lee C Potter
- Randolph L. Moses
Organizations
- Ohio State University