Algorithm development for buried object detection and identification using ground penetrating radar

Improvised explosive devices (IEDs) have recently become a considerable concern for governments' national security due to increased usage. Therefore, detecting IEDs is a significant problem now. Since IEDs do not have a particular shape, detecting and classifying IEDs becomes a complex problem. Due to the irregularity of wire shape, recent research focuses on detecting command wire, which is the triggering mechanism of IEDs. The study in this paper proposes a detection and clustering algorithm for the wire detection problem. The synthetic data were generated using the gprMax software, an opensource FDTD simulation environment. Different wire and clutter orientations were simulated while creating a 2-D GPR database. Three different prescreening algorithms were compared concerning computational time and Signal to Noise ratio. The Go-Decomposition method was used at the preprocessing stage. The discrimination wire from clutter has been conducted using the K-Means clustering method. The proposed algorithm results show promising outcomes over simulated GPR 2- D C-scans.