Nonparametric mine line detection using spatial analysis

摘要

A mine line detection method to identify mine-like contacts (MLCs) that fit into mine line patterns dictated by standard mine laying protocol. In general, MLCs define a minefield where the minefield has both mines and false alarms. The method classifies every MLC into one of three classes: fixed-a mine line pattern with fixed mine spacing; random-a mine line pattern with random mine spacing; or false alarm-the MLC does not fit into a mine line pattern and is not considered a mine. The method requires no knowledge of either the sensor probability of detection (Pd) or probability of false alarm (Pfa); nonetheless, it significantly improves both. A nonparametric anomaly detector identifies regions with unusually high MLC density for more detailed analysis. Suspected regions are analyzed for unusually regular patterns of MLCs. Probabilities that the observed target densities and target patterns could occur by chance are estimated.