Demonstration of Physics-Inspired Classification Methodologies for Munitions Response
Dr. Dean Keiswetter | Acorn SI
Objectives of the Demonstration
SAIC and NAEVA Geophysics have partnered to demonstrate the classification capabilities of combining broadband electromagnetic induction (EMI) data collected using the latest advanced EMI sensing systems with physics-based inversion routines made available in the UX-Analyze module of Oasis Montaj. By partnering a research firm with a production geophysical services contractor, the goal is to promote the adoption, understanding, and acceptance of classification approaches designed for munitions remediation. If successful, this project will transfer technology developed by SERDP and ESTCP to the production community, which can greatly increase the cost-efficiency of production-level cleanup.
This project will demonstrate and document the capabilities of UX-Analyze to successfully classify electromagnetic signatures acquired by advanced sensors as munitions or clutter. UX-Analyze includes procedures for isolating and culling target signatures from reconnaissance EMI data or using cued data collected by advanced sensors to assign class labels using library matching procedures or statistical classifiers. The analysis routines will be documented through individual anomaly plots, target databases, and processing logs. The project will establish a baseline of what can be accomplished with standard EM61 survey data, determine what performance improvements can be achieved using advanced EMI sensors, and document how performance varies among the sites investigated.