Detection and identification of landmines using electromagnetic induction spectroscopy

Reducing the false alarm rate is one of the most outstanding problems in clearing landmines. A broadband EMI sensor combined with a spectral matching method, known as Electromagnetic Induction Spectroscopy (EMIS), has shown great potential to address this complex problem. The common EMIS discrimination approach is to match the spectrum of an unknown target to a library of known spectra. A broadband EM sensor works like a metal detector, but it can operate at multiple programmable frequencies suitable to a given geologic environment. Once it detects a potential target, the sensor will interrogate the target and measure its spectral responses over the entire operating bandwidth. The sensor will then compare the measured spectrum with a library of spectra stored for mines that are known or presumed to occur in the area. The process will generate a rank-ordered list of spectral matches. We have developed and field-tested several spectral matching algorithms for detecting and identifying buried landmines. Recent EMIS test surveys at a controlled military site have produced a high detection rate at a very low false alarm rate. This paper will present results from the surveys including the spectral matching algorithms employed for the identification process.