Depth and Radius Joint Estimation for Underground Pipeline Using GPR

Qimin Xu,Zuoliang Wang,Xu Li

DOI: https://doi.org/10.1109/jsen.2023.3280177

IF: 4.3

2023-01-01

IEEE Sensors Journal

Abstract:In this article, a deep learning-based methodology is proposed to estimate depth and radius for underground pipeline using ground-penetrating radar (GPR) images. The proposed methodology conducts the depth and radius joint estimation based on GPR B-scan image preprocessing. During the preprocessing, the autoencoder structure network is designed to binarize the GPR image. Besides, redundant detection boxes of hyperbola detection are also eliminated. For pipeline depth estimation, the parabola model is used instead of the hyperbola model to obtain the symmetry axis of the pipeline GRP image more efficiently. The consecutive a-scan data at extracted symmetry axis and its vicinity are spliced and sent to the long short-term memory (LSTM) network for depth estimation. For pipeline radius estimation, the image features and depth information are concatenated. A pretrained convolutional neural network (CNN) is designed to extract the GPR image features. Meanwhile, a 2-D-Mapping network is proposed to encode the 1-D depth information into 2-D feature matrix. Improving the dimension of depth information rather than flattening the GPR image features (dimension reduction) can preserve the spatial information of image features. Besides, the depth information can be concatenated with the image features more flexibly. In all, the proposed methodology can accurately estimate the depth and radius of underground pipeline. To verify the effectiveness of the proposed methodology, the gprMax is used to generate the pipeline GPR images in inhomogeneous soil and create the dataset. The experiments include detailed comparison and analysis, illustrating that the proposed methodology is feasible and effective.