Detection, Tracking and Classification of Multiple Targets using Advanced Beamforming and Classification Methods
Abstract
Detection, tracking and localization of ground targets, e.g. trucks and tanks, using unattended passive acoustic sensor arrays is considered. Several factors complicate this problem. These include: non-stationary target signatures, loss of coherence and fading, wind noise and weatherlenvironmental effects, competing clutter, lack of any prior knowledge about the number of targets, and presence of multiple targets in tight formations. The goal of this Phase I research was to develop fast and accurate direction of arrival (DOA) estimation and source separation algorithms to localize and resolve multiple closely spaced targets. The array geometry was considered to develop two classes of algorithms. The first class is based upon high resolution subspace-based DOA estimation methods that can be used to localize and track the targets. The second class, however, is based upon source separation methods using variants of independent component analysis for determining the number of potential targets. Of particular importance was to investigate target detection and information extraction in spatially aliased data due to the sparse nature of the array. Testing and benchmarking of the algorithms was done on real and synthesized data sets containing multiple target scenarios in abreast, staggered, or single-file formations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 28, 2002
- Accession Number
- ADA409366
Entities
People
- K. Khorasani
- M. R. Azimi-sadjadi