A Probabilistic Target Classification and Description Model for Seismic Sensors.
Abstract
Unattended ground sensors have a tremendous potential for providing information about battlefield targets, but for the most part this potential has been unrealized. The Marine Corps has recently fielded the Phase V seismic sensors of the Tactical Remote Sensor System (TRSS). These sensors are more sensitive than any of the previous versions, and their potential to provide more detailed target information is also greater than that of previous sensors. The current target classification and description model used by TRSS was developed for sensors which were placed in use in the early 1960's. The model is simple and deterministic in nature, and does not take into account the variance in the sensor system or the variance in sensor performance due to target type, target velocity, soil composition, or other potential factors. This thesis examines the sensor system variance and the effect of target type on sensor performance through field testing and develops an improved model for target description that accounts for these effects. The revised model takes advantage of the measured sensor characteristics to better describe the target, and provides the user with bounds that describe the credibility of the model's estimate.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1994
- Accession Number
- ADA288548
Entities
People
- Mark D. Van Kan
Organizations
- Naval Postgraduate School