LAUSR

Improvised explosive devices (IEDs) pose a significant threat to defense forces and humanitarian demining personnel. They are weapons of modern times, made from nonconventional military materials, rendering them difficult to identify when buried in the ground. Numerous studies focus on detecting these explosive threats and reducing the false alarm rate. However, there are few attempts to identify the detected explosive devices to take proper countermeasures. This article presents a multilevel projective dictionary learning (DL) method to classify ground-penetrating radar signals from IEDs. The proposed dictionary learning method solves three different tasks simultaneously: suppressing background clutter, learning a set of discriminative features for classification, and training a classifier. The suppression of ground clutter is formulated as a low-rank (LR) optimization problem with sparse constraints, where a low-rank subspace is learned from background clutter signals. Dictionary learning is used to transform the target signals into discriminative feature vectors, which are in turn used by the classifier to predict the target class. Experiments were conducted on real radar data. The results showed that the proposed method is more effective than the existing dictionary models and machine learning methods.