FilterNet: A CNN-RNN based filter model used for raw tunnel lining GPR data

Abstract

Ground-Penetrating Radar (GPR) technology, with its characteristics of being fast, non-destructive, and high-resolution, has become an important tool for detecting underground structures. However, GPR data inevitably suffer from environmental noise and electromagnetic interference during the acquisition process, leading to decreased data quality and increased complexity in data processing. Traditional filtering algorithms have limitations such as low discrimination between noise and signal, poor adaptability, and inability to process data in real time. This paper proposes a filtering model based on deep neural networks, called FilterNet. FilterNet combines Convolution Neural Networks (CNN) and recurrent neural networks (RNN) for processing multi-channel data. It can perform end-to-end filtering directly on the raw tunnel lining GPR data, achieving functions such as removing air reflection waves, denoising, and automatic gain. Using PSNR (Peak Signal-to-Noise Ratio) and SSIM (Structural Similarity Index) as statistical indicators, it is shown that the FilterNet model improves filtering precision. The SSIM of all three models is 0.997, and the PSNR of FilterNet1D and FilterNet are 19.06 and 19.41, respectively. Furthermore, tests on the model's processing efficiency indicate that FilterNet requires less memory and is more efficient than the UNet model. FilterNet's parameters are only 48 ​% of those of UNet. Its GFLOPS (Giga Floating Point Operations Per Second) is only one-third of UNet's, and it can process data in real time. Additionally, FilterNet performs exceptionally well in suppressing random noise.

Keywords

Ground penetrating radar / Deep learning / Data filtering / Recurrent neural network / Convolution neural network

CRediT authorship contribution statement

Bang Zhang: Writing - review & editing, Writing - original draft, Resources, Methodology. Yu-Qi Cai: Writing - review & editing, Writing - original draft, Formal analysis. Zi-Ye Yu: Writing - review & editing, Formal analysis, Data curation, Conceptualization. Kai Li: Writing - review & editing, Data curation.

Author agreement and acknowledgments

All authors agree for this publication. The authors thank the anonymous reviewers for the constructive suggestions and the editor for their help during the submission. This work is supported by China Railway Construction Corporation Limited (CRCC) Major Science and Technology Special Project: Research, Development, and Application Demonstraction of Gelogical Information Digital Platform (Project No.: 2021-A02).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

[1] Cheng, Q., Cui, F., Chen, B., Dong, G., Wang, R., Zhang, G., Fu, J., 2024. Attenuation of non-stationary random noise in ground penetrating radar data based on time-varying filtering. Measurement 236, 115169.

[2] Chicarella, S., Ferrara, V., D'Atanasio, P., Frezza, F., Pajewski, L., Pavoncello, S., et al., 2014. Analyses and measures of GPR signal with superimposed noise. In: EGU General Assembly Conference Abstracts, p. 5387.

[3] Cho, K., 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.

[4] He, W., Hao, T., Ke, H., Zheng, W., Lin, K., 2020. Joint time-frequency analysis of ground penetrating radar data based on variational mode decomposition. J. Appl. Geophys. 181, 104146.

[5] Huang, Y., Zhou, W., 2023. Ground penetrating radar image de-noising method based on multi-noise and self-supervised learning. In: 2023 5th International Conference on Intelligent Control, Measurement and Signal Processing (ICMSP). IEEE, pp. 970 975.

[6] LeCun, Y., Bottou, L., Bengio, Y., Haffner, P., 1998. Gradient-based learning applied to document recognition. Proc. IEEE 86 (11), 2278

[7] Li, R., Zhang, H., Chen, Z., Yu, N., Kong, W., Li, T., et al., 2022. Denoising method of ground-penetrating radar signal based on independent component analysis with multifractal spectrum. Measurement 192, 110886.

[8] Lin, H., Xiao, J., Liu, Z., Liu, Z., Deng, Y., 2023. Clutters suppression in GPR signal for railway subgrade detection based on deep learning. Prog. Geophys. 38 (6), 2714 2723.

[9] Liu, H., Wang, S., Jing, G., Yu, Z., Yang, J., Zhang, Y., Guo, Y., 2023. Combined CNN and RNN neural networks for GPR detection of railway subgrade diseases. Sensors 23(12), 5383.

[10] Liu, Z., Xiao, J., Shen, R., Liu, J., Guo, Z., 2024. Deep learning-based suppression of strong noise in GPR data for railway subgrade detection. IEEE Trans. Geosci. Rem. Sens.

[11] Oskooi, B., Parnow, S., Smirnov, M., Varfinezhad, R., Yari, M., 2018. Attenuation of random noise in GPR data by image processing. Arabian J. Geosci. 11, 1 10.

[12] Rohman, B.P., Nishimoto, M., Ogata, K., 2021. Reconstruction of missing ground- penetrating radar traces using simplified U-Net. Geosci. Rem. Sens. Lett. IEEE 19, 1-5

[13] Ronneberger, O., Fischer, P., Brox, T.,2015. U-net:convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III, vol. 18. Springer International Publishing, pp. 234 241.

[14] Temlioglu, E., Erer, I., 2021. A novel convolutional autoencoder-based clutter removal method for buried threat detection in ground-penetrating radar. IEEE Trans. Geosci. Rem. Sens. 60, 1 13.

[15] Xu, X., Lei, Y., Yang, F., 2018. Railway subgrade defect automatic recognition method based on improved faster R-CNN. Sci. Program. 2018 (1), 4832972.

[16] Xue, W., Dai, X., Zhu, J., Luo, Y., Yang, Y., 2019. A noise suppression method of ground penetrating radar based on EEMD and permutation entropy. Geosci. Rem. Sens. Lett. IEEE 16 (10), 1625 1629.

[17] Yang, S.S., Liu, C., Li, G., Li, Y., 2023. Review of data processing methods for ground penetrating radar. Hans J. Civ. Eng. 12, 25.

[18] Zhang, S.W., Wu, R.X., Han, Z.A., et al., 2021. The application of bilateral filtering to denoise processing of ground penetrating radar data. Geophys. Geochem. Explor. 45 (2), 496 501, 2021.

[19] Zhu, J., Xue, Y., Zhang, N., Li, Z., Tao, Y., Qiu, D., 2017. A noise reduction method for Ground Penetrating Radar signal based on wavelet transform and application in tunnel lining. In: IOP Conference Series: Earth and Environmental Science, vol. 61. IOP Publishing, 012088, 1.