Wide Area Detection and Mapping Technologies for Locating Minefields Containing Anti-Personnel Landmines

Period of Performance: 05/13/1999 - 12/13/1999

$99.1K

Phase 1 SBIR

Recipient Firm

Arcon Corp.
260 Bear Hill Road
Waltham, MA 02451
Principal Investigator

Abstract

We describe a new approach for identification and location of buried non-metallic mines from backscatter data from a wide-footprint microwave source. A method based on the Feynman diagram model is described that permits discrimination between multiply scattered field data and data arising fro discrete scatterers such as mines. The limited backscatter data associated with discrete but multiply scattering objects is then processed using a nonlinear filter to remove multiple scattering effects. The resulting poorly resolved estimate for each discrete object is further processed using our spectral estimation technique, the Prior Discrete Fourier Transform (PDFT). This technique requires that some prior knowledge about the targets, i.e. size, approximate shape. Using this PDFT information allows discrimination between targets based on a simple calculation of the magnitude of the coefficients of the PDFT estimate derived from each possible choice of prior knowledge. In Phase I we propose to develop the first version of a complete software package enabling mine detection based on these models. We will conduct trials of the effectiveness of our approach applied to simulated and real data. We will specify a complete system, including the required radar hardware and signal processing capability, to be constructed and tested in Phase II. The proposed real-time microwave imaging algorithm is not only useful for the mine detection application, but also for application in : (1) medical imaging, real-time monitoring of tumor growth where the microwave or higher frequencies that would be employed provide a non-intrusive, innocuous means for tumor detection and monitoring; and (2) industrial applications such as non-destructive testing (crack detection and dislocation identification) in material structures. A further application of our imaging algorithm might be to provide an imaging training and teaching device for use in academic institutions in both the medical imaging and engineering fields.