A Complex Approach to UXO Discrimination: Combining Advanced EMI Forward Models and Statistical Signal Processing

Abstract

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.

Demonstration of Advanced EMI Models for Live-Site UXO Discrimination at Former Camp Butner, North Carolina

“…The advanced EMI models we present here (ONVMS and JD) were developed under SERDP Project MM-1572 [5] and tested against TEMTADS data sets collected at the Aberdeen Proving Ground (APG) test site in Maryland and in a retrospective analysis of cued TEMTADS data collected at the San Luis Obispo site [6][7][8][9][10]. The present test of discrimination performance considers data taken at Camp Butner; this live site was densely contaminated with small targets such as 37-mm projectiles and fuzes, adding another level of complexity into the classification and thus further demonstrating the robustness of the advanced EMI models for live-site UXO discrimination.…”

“…The advanced EMI models and statistical signal processing approaches developed and tested over the past three years under SERDP Project MM-1572 [5] were able to detect and identify buried UXO ranging in caliber from 25 mm up to 155 mm. The technique was seen to be physically complete, fast, accurate, and clutter-tolerant, and provided excellent classification in both single-and multiple-target scenarios when combined with multi-axis/transmitter/receiver sensors like TEMTADS and MetalMapper [11].…”

“…The sensor thus provides 25  25 spatial data points at any given time channel t q , q = 1, 2,…, N q . If we define H k,m as the z-component of the magnetic field measured by the m-th receiver coil when the k-th transmitter is active, then each row of the measured multi-static response (MSR) data matrix 5 April 2012…”

“…Inverse-scattering problems are solved by defining an objective function [5] that measures the mismatch between modeled and measured magnetic-field data and proceeding to find its global minimum. Standard gradient search approaches often suffer from a profusion of local minima that sometimes result in incorrect location and orientation estimates.…”

“…To avoid this problem we recently adapted and incorporated a different class of global optimization search algorithms, among them differential evolution (DE) [19][20], a heuristic, parallel, direct-search method for minimizing non-linear functions of continuous variables that is straightforward to implement and has good convergence properties. We have combined DE with ONVMS to invert digital geophysical EMI data [5,14]. All EMI optimizations are split into linear and nonlinear parts, alternating between the two and iterating to minimize the objective function.…”