Fort Ord Site UXO Classification Demonstration using Fully Polarimetric GPR
Abstract
The OSU/ESL broadband, full-polarimetric ground penetrating radar (GPR) has been applied for UXO classification under the support of ESTCP project beginning in 1999. Since then the system has been used to collect radar data and perform blind UXO classification at several UXO sites including Tyndall AFB, FL ("Tyndall"), Blossom Point, MD (BP) and Jefferson Proving Ground, IN (JPG). The results of these field tests have been documented in previous ESTCP reports. The fourth and final field demonstration was conducted at the former Fort Ord, CA, from October 29 to November 3 of 2001. The first blind classification results were then reported to ESTCP prior to the reception of the true depth information. This report presents documentation of the test and the results of classification processing, with and without that depth information. The objective of the Fort Ord test is to evaluate classification performance in a sandy environment, given the improvements in survey and processing systems developed since the Tyndall demo. Detailed site information for the Fort Ord UXO site can be found in the Ordnance Detection and Discrimination Study (2000) prepared by the USA Environmental, Inc. Although the earlier Tyndall site was also a sandy site, that demo was performed more then three years ago with processing based only on single-position features. The improved classification algorithm utilizes late-time radar signatures including natural resonance [1][2] and detection of linear scattering polarization, as introduced during the Tyndall test [3]; and analysis of the spatial variation of data features as introduced during the BP demo [4][5]. Multiple radar passes were also performed for each target spot. The final system combined multiple positions, multiple orientations, broad frequency range and fully polarimetric configuration.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 2005
- Accession Number
- ADA606889
Entities
People
- Chi-chih Chen
- Kevin O’neill
Organizations
- Cold Regions Research and Engineering Laboratory