Clutter removal for landmine using different signal processing techniques

This paper demonstrates the clutter reduction using a stepped-frequency ground penetrating radar data for anti-personal landmines detection. For the purpose of clutter reduction, five methods have been studied and applied to experimental data. And their output SNR have been compared. These methods are simple mean subtraction, likelihood ratio test, system identification, subspace singular value decomposition technique and wavelet transform. The experimental data have been collected by using frequency stepped GPR system covering the range from 1.5 GHz to 4 GHic. The results demonstrate that the wavelet transform outperforms the other techniques for clutter reduction and gives the best SNR.