Statistical and Adaptive Signal Processing For UXO Discrimination for Next-Generation Sensor Data

Abstract

Until recently, detection algorithms could not reliably distinguish between buried UXO and clutter, leading to many false alarms. Over the last several years modern geophysical techniques have been developed that merge more sophisticated sensors, underlying physical models, statistical signal processing algorithms, and adaptive training techniques. These new approaches have dramatically reduced false alarm rates, although for the most part they have been applied to data collected at sites with relatively benign topology and anomaly densities. On more challenging sites, performance of even these more modern discrimination approaches is still quite poor. As a result, efforts are underway to develop a new generation of UXO sensors that will produce data streams of multi-axis vector or gradiometric measurements, for which optimal processing has not yet been carefully considered or developed. We describe a research program to address this processing gap, employing a synergistic use of advanced phenomenologicalmodeling and signal-processing algorithms. The key foci of the program are (1) development of new physics-based signal processing approaches applicable to the problem in which vector data is available from such sensors; and (2) development of the theory of optimal experiments to guide the optimal design and deployment of the new sensor modalities. Here, we present initial results using simulated data obtained with our phenomenological models that indicate that optimal processing of features extracted from multi-axis EMI data can provide substantial improvements in discrimination performance over processing of features extracted from single-axis data.