Radar Detection,Discrimination,and Classification of Buried Non-metallic Mines. Volume 1.

摘要

A program for development of an off-road mine detection system was tasked to gather and analyze wideband radar spectral data to determine detection and classification potential for buried objects of interest. In particular, the portion of the spectrum which is most useful in the detection, discrimination, and classification of buried targets was to be identified. Two sets of data were taken. Wideband CW radar data were collected using a computer-driven automatic network analyzer. Wideband pulse radar data were collected by the Georgia Institute of Technology Engineering Experiment Station using a Government Furnished Equipment (GFE) short-pulse radar owned by MERADCOM. The wideband target signatures obtained using the MERADCOM short-pulse radar were used to verify the utility of the processing algorithms which were developed from wideband CW radar data. The performance of each processing step--detection, discrimination, and classification--was acceptable for the specified scenario when the soil was relatively dry (i.e., zero to ten percent moisture content). However, when the soil was more moist than this (e.g., ten to thirty percent), the performance of the processor deteriorated noticeably.