Translator Disclaimer
16 April 2008 Segregation of tracked and wheeled ground vehicle mobility mechanisms through in-situ adaptation of seismic features
Author Affiliations +
Abstract
The ability to perform generalized ground vehicle classification by unattended ground sensors (UGS) is an important facet of data analysis performed by modern unattended sensor systems. Large variation in seismic signature propagation from one location to another renders exploiting seismic measurements to classify vehicles a significant challenge. This paper presents the results of using an adaptive methodology to distinguish between tracked and wheeled ground vehicle mobility mechanisms. The methodology is a passive in-situ learning process that does not rely upon an explicit calibration process but does require an estimated range to the target. Furthermore, the benefits of the seismic feature adaptation are realized with a sparse information set. There exist scenarios in which the adaptation fails to provide information when implemented as an independent process. These situations, however, may be mitigated by sharing information with other classification algorithms. Once properly initialized, the in-situ adaptation process correctly categorizes over 95% of ground vehicles.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Christopher G. Park, James Fitzgerald, and Dennis Power "Segregation of tracked and wheeled ground vehicle mobility mechanisms through in-situ adaptation of seismic features", Proc. SPIE 6963, Unattended Ground, Sea, and Air Sensor Technologies and Applications X, 69630X (16 April 2008); https://doi.org/10.1117/12.784435
PROCEEDINGS
8 PAGES


SHARE
Advertisement
Advertisement
Back to Top