Signal Processing of Radar Signatures Using Hidden Markov Model and Its Application to Subsurface Detection

摘要

Signal processing of sequential Ground Penetrating Radar (GPR) signatures for detecting and identifying buried objects is studied. Classifiers are designed by using hidden Markov models (HMMs) to characterize the sequential information in multi-observation GPR signatures. Detection and classification results are presented for some landmine-like targets in underground. The example results show that good detection and classification performances are obtained, although the targets are very similar to each other.