Wavelet-based method for detection of shallowly buried objects fromGPR data

摘要

A method for simultaneous clutter suppression and signal detectionin ground penetrating radar (GPR) data is presented. It computes arunning estimate of the background signal and detects abrupt changesfrom the estimated signal, assumed to correspond to the returns from aburied object, using a translation invariant (TI) wavelet packetdecomposition. The detection statistic is defined as the sum of severalpeak energy values of the TI wavelet representation and is compared toan adaptively determined threshold. The technique performedsatisfactorily when tested on a number of targets buried in severaldifferent soil environments