Traditional single-polarimetric ground penetrating radar (GPR) has been widely used for the detection and classification of cavities, pipes, cables, and other subsurface objects. However, as these subsurface objects show similar hyperbolic patterns in B-scan images, their data interpretation remains a challenge. In this letter, we propose an underground target classification method based on full-polarimetric GPR data and a multi-branch deep convolutional neural network (CNN) with channel attention modules (CAMs). Here, the full-polarimetric GPR data as input is separately sent into each branch of the network for feature extraction. The CAM is used to improve the network’s sensitivity and processing of key features. After fusing these full-polarimetric features, underground target classification is achieved with high accuracy. Our experiments demonstrate that more target information can be obtained by applying full-polarimetric GPR data, which is beneficial for improving target classification accuracy.

Ground-penetrating radar (GPR) is an effective tool for underground object detection, but its data interpretation remains a great challenge. In this letter, we propose a novel underground object classification algorithm using deep 3-D convolutional networks (C3D) and multiple mirror encoding (MME) for 3-D GPR data. Although deep learning technique has been applied to interpret the GPR data, most of the existing methods are based on GPR B-scans and have a relatively low accuracy since the reflections from various subsurface targets present similar hyperbolic patterns in B-scans. To improve the classification accuracy, we use 3-D GPR data as training data for C3D to capture the spatio-temporal features between parallel B-scans. Since 3-D GPR data including single object has different sizes in consideration of actual sizes of objects, they are rearranged by the MME method to enhance spatio-temporal features, as well as to satisfy the requirement of the network input. Experimental results demonstrate that the proposed method outperforms the state-of-the-art B-scan-based methods.

A correlation weighted back projection (CWBP) imaging algorithm is proposed to suppress artifacts introduced by the classic back projection (BP) algorithm in the application of chaotic ground-penetrating radar (GPR). In the proposed CWBP algorithm, the backscattering response to each point in the imaging region is first reconstructed through the weighting, which is used to highlight the backscattering response of targets. Then the quadratic cross correlation is calculated to reduce the nontargets’ backscattering. The experimental results show that the artifacts are significantly suppressed. In addition, the ghosts generated by overlapped of multitargets artifacts in the classic BP are removed. Our results also show that the proposed CWBP can be applied for the stepped frequency continuous wave radar.

Clutter removal is of vital importance to underground detection via ground penetrating radar (GPR) since the target response is usually overwhelmed by strong clutter. The low-rank and sparse decomposition (LRSD) method, which decomposes the GPR data into a low-rank background matrix and a sparse target matrix, has been applied in GPR. However, the existing LRSD-based methods, such as robust principal component analysis (RPCA) and robust non-negative matrix factorization, are sensitive to the regularization parameter or have high computation complexity. In this letter, a novel clutter removal based on factor group-sparse regularization is proposed. It uses a nonconvex matrix factorization as a surrogate for the matrix rank rather than the nuclear norm as in the RPCA. It has good efficiency and robustness to the parameters. Simulation and experimental results demonstrate that the proposed method has higher performance in terms of the improvement factor values and lower computational complexity than the state-of-the-art methods.