Target Detection in SFCW Ground Penetrating Radar with C3 Algorithm and Hough Transform based on GPRMAX Simulation and Experimental Data

Abstract

We perform target detection in B-scan images obtained using stepped frequency continuous wave (SFCW) ground-penetrating radar (GPR) simulation with Matlab and the open-source software for electromagnetic propagation simulation gprMax and using experimental data from measurements available on the Web. We perform hyperbola detection and target depth estimation using the Hough transform. Additionally, we reduce the complexity of Hough transform using the Column-Connection Clustering (C3) algorithm. The results show that gprMax can be used with SFCW modulated signals, besides its use with conventional impulse signals (impulse GPR). The proposed combination of C3 and Hough transform shows very good detection capability. These algorithms are also tested on experimental GPR data available on the Web to show their applicability in real soil conditions. Mean subtraction method and Singular Value Decomposition (SVD) algorithms are used to reduce the effects of the ground reflected wave.