High-efficiency and high-quality detection of oil pipeline will significantly reduce environmental pollution and economic loss, so an unconventional oil pipeline anomaly detection convolutional neural network (CNN) algorithm based on attention mechanism is proposed in this article. By taking the simulated ground-penetrating radar (GPR) data as prior knowledge, the structure of the convolutional neural network based on the attention mechanism is constructed, and finally, the location and working condition of the underground oil pipeline are recognized in the simulation data and measured data. The simulation results show that after using the new optimized convolutional neural network, the accuracy rates of the leakage discrimination from horizontal data acquired along the oil pipeline and the classification of the target from longitudinal data acquired perpendicular to the oil pipeline are 94.5% and 84.6%, respectively. Compared with the original convolutional neural network without an attention mechanism, the accuracy rates of the leakage discrimination and the classification of the target are improved by 6.2% and 7.8%, respectively. We further train measured data with an optimized convolutional neural network, results show that compared with a conventional network, the new network can increase the corresponding accuracy rates of the leakage discrimination and the targets classification by 5.4% and 6.9%, reaching 92.3% and 84.4%, respectively. According to our study, the ground-penetrating radar oil pipeline recognition algorithm based on an attention mechanism can well accomplish the identification of underground oil pipelines.

To address the significant impact of noise on the target detection performance of borehole radar (BHR), a key type of ground-penetrating radar (GPR), a denoising scheme based on the whale optimization algorithm (WOA) for adaptive variational mode decomposition (VMD) and multiscale principal component analysis (MSPCA) is proposed. This study initially conducts the modal decomposition of BHR data using an improved adaptive VMD method based on the WOA; it then automatically selects modes meeting specific frequency band standards. The correlation coefficients between these modes and the original signal are computed, discarding weakly correlated modes before signal reconstruction. Finally, MSPCA further suppresses noise, yielding denoised BHR data. Simulations show that the proposed scheme increases the signal-to-noise ratio by 17.964 dB or higher, surpassing the more established denoising techniques of robust principal component analysis (RPCA), MSPCA, and empirical mode decomposition (EMD), and obtains the most favorable results in terms of the RMSE and MSE metrics. The experimental results demonstrate that the proposed scheme more effectively suppresses vertical and random noise signals in BHR data. Both the numerical simulations and experimental results confirm the effectiveness of this scheme in noise reduction for BHR data.

This contribution deals with the focusing and time–depth conversion of GPR data in a two-layer scenario. In a layered scenario, the different propagation velocities of the electromagnetic waves in different layers can induce distortion in the reconstruction. In particular, it is not possible, with common data processing, to achieve the focusing of both the targets embedded in the upper and deeper layers at the same time. We will present a strategy for mitigating this problem. The strength points of this strategy are its simplicity and the fact that it only requires a commercial code for the GPR data processing plus some steps easily implementable through a homemade code. This work extends a contribution of the same authors at the 20th International GPR conference held in Chanchung, China, in June 2024, and the conclusions are updated with three-dimensional numerical results, including depth slices determined directly in the spatial domain. Both numerical and experimental data will be shown.

This paper addresses the problem of retrieving the propagation velocities of electromagnetic waves in a non‐homogeneous soil made up of two layers separated by a not‐flat interface. The propagation velocities are estimated from GPR data gathered at the air–soil interface in common offset configuration. The well‐known diffraction hyperbola method is exploited for retrieving the propagation velocity in the shallower layer. Conversely, an extension of the diffraction hyperbola method is proposed to estimate the propagation velocity in the second layer (meant as the deeper one). The proposed approach accounts numerically for the refraction of the waves at the buried interface, which deforms the diffraction curve. The shape of the buried interface is accounted for by using abscissa‐time samples picked up from the data. The approach exploits theoretical aspects derived for a flat buried interface and empirically extends them to the case of a curved (smooth) buried interface. Experimental results in a controlled scenario are presented.