Deep Learning-Based Ground-Penetrating Radar Inversion for Tree Roots in Heterogeneous Soil
Abstract
:1. Introduction
2. Materials and Methods
2.1. Overall Architecture
2.2. Pyramidal Convolution Feature Extraction Block
2.3. Vision Transformer Feature Extraction Block
3. Experiments and Analysis of Synthetic Data
3.1. Building the Synthetic Dataset
3.1.1. Overview of Simulation Model
3.1.2. Layered Heterogeneous Soil and Buried Objects
3.2. Simulation Data Preprocessing
3.3. Comparison with Other Methods
3.3.1. Loss Function and Evaluation Indicators
3.3.2. Experimental Environment and Parameter Setting
3.3.3. Inversion Results of Simulation Data
3.4. Ablation Experiment
4. Experiments and Analysis of Real Data
4.1. Real Dataset and Experimental Implementation
4.2. Analysis of the Experimental Results
5. Conclusions and Perspectives
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Alani, A.M.; Lantini, L. Recent advances in tree root mapping and assessment using non-destructive testing methods: A focus on ground penetrating radar. Surv. Geophys. 2020, 41, 605–646. [Google Scholar] [CrossRef]
- Hugenschmidt, J.; Kay, S. Unmasking adaption of tree root structure in agroforestry systems in switzerland using gpr. Geoderma Reg. 2023, 34, e00659. [Google Scholar] [CrossRef]
- Lantini, L.; Alani, A.M.; Giannakis, I.; Benedetto, A.; Tosti, F. Application of ground penetrating radar for mapping tree root system architecture and mass density of street trees. Adv. Transp. Stud. 2019, 3, 51–62. [Google Scholar]
- Sun, D.; Jian, F.; Wu, H. Root location and root diameter estimation of trees based on deep learning and ground-penetrating radar. Agronomy 2023, 13, 344. [Google Scholar] [CrossRef]
- Dou, Q.; Wei, L.; Magee, D.R.; Cohn, A.G. Real-time hyperbola recognition and fitting in gpr data. IEEE Trans. Geosci. Remote. Sens. 2016, 55, 51–62. [Google Scholar] [CrossRef]
- Lai, W.W.L.; Derobert, X.; Annan, P. A review of ground penetrating radar application in civil engineering: A 30-year journey from locating and testing to imaging and diagnosis. Ndt E Int. 2018, 96, 58–78. [Google Scholar]
- Cheng, X.; Shao, W.; Wang, K.; Wang, B.Z. Uncertainty analysis in dispersive and lossy media for ground-penetrating radar modeling. IEEE Antennas Wirel. Propag. Lett. 2019, 18, 1931–1935. [Google Scholar] [CrossRef]
- Hahm, J.; Kim, S.D.; Kim, D.; Choi, S.H. Frequency-shifting-based b2d-dct-pca for handheld landmine detection using gpr data. IEEE Geosci. Remote Sens. Lett. 2020, 18, 1560–1564. [Google Scholar] [CrossRef]
- Serhir, M.; Lesselier, D. (Eds.) Ground Penetrating Radar: From Theoretical Endeavors to Computational Electromagnetics, Signal Processing, Antenna Design and Field Applications; John Wiley & Sons: Hoboken, NJ, USA, 2024. [Google Scholar]
- Kuroda, S.; Takeuchi, M.; Kim, H.J. Full-waveform inversion algorithm for interpreting crosshole radar data: A theoretical approach. Geosci. J. 2007, 11, 211–217. [Google Scholar] [CrossRef]
- Travassos, X.L.; Avila, S.L.; Ida, N. Artificial neural networks and machine learning techniques applied to ground penetrating radar: A review. Appl. Comput. Inform. 2021, 17, 296–308. [Google Scholar] [CrossRef]
- Ronneberger, O.; Fischer, P.; Brox, T. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention—MICCAI 2015: 18th International Conference, Munich, Germany, 5–9 October 2015; Proceedings, Part III; Springer International Publishing: Cham, Switzerland, 2015. [Google Scholar]
- Liu, B.; Ren, Y.; Liu, H.; Xu, H.; Wang, Z.; Cohn, A.G.; Jiang, P. Gprinvnet: Deep learning-based ground-penetrating radar data inversion for tunnel linings. IEEE Trans. Geosci. Remote Sens. 2021, 59, 8305–8325. [Google Scholar] [CrossRef]
- Dai, Q.; Lee, Y.H.; Sun, H.H.; Ow, G.; Yusof, M.L.M.; Yucel, A.C. Dmrf-unet: A two-stage deep learning scheme for gpr data inversion under heterogeneous soil conditions. IEEE Trans. Antennas Propag. 2022, 70, 6313–6328. [Google Scholar] [CrossRef]
- Ge, J.; Sun, H.; Shao, W.; Liu, D.; Yao, Y.; Zhang, Y.; Liu, R.; Liu, S. GPR-TransUNet: An improved TransUNet based on self-attention mechanism for ground penetrating radar inversion. J. Appl. Geophys. 2024, 222, 105333. [Google Scholar] [CrossRef]
- Zhang, K.; Cheng, X.; Zhao, Y. Investigation of ground-penetrating radar inversion methods in heterogeneous soil conditions. Chin. J. Eng. Geophys. 2023, 20, 843–854. [Google Scholar]
- Svoboda, P.; Haberle, J.; Moulik, M. The effect of drip irrigation on the length and distribution of apple tree roots. Horticulturae 2023, 9, 405. [Google Scholar] [CrossRef]
- Ioffe, S. Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv 2015, arXiv:1502.03167. [Google Scholar]
- Daubechies, I.; DeVore, R.; Foucart, S.; Hanin, B.; Petrova, G. Nonlinear approximation and (deep) relu networks. Constr. Approx. 2022, 55, 127–172. [Google Scholar] [CrossRef]
- Li, C.; Fan, Y.; Cai, X. Pyconvu-net: A lightweight and multiscale network for biomedical image segmentation. BMC Bioinform. 2021, 22, 14. [Google Scholar] [CrossRef]
- Woo, S.; Park, J.; Lee, J.Y.; Kweon, I.S. Cbam: Convolutional block attention module. In Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany, 8–14 September 2018; pp. 3–19. [Google Scholar]
- Yuan, L.; Chen, Y.; Wang, T.; Yu, W.; Shi, Y.; Jiang, Z.H.; Yan, S. Tokens-to-token vit: Training vision transformers from scratch on imagenet. In Proceedings of the IEEE/CVF International Conference on Computer Vision, Virtual, 11–17 October 2021; pp. 558–567. [Google Scholar]
- Warren, C.; Giannopoulos, A.; Giannakis, I. gprmax: Open source software to simulate electromagnetic wave propagation for ground penetrating radar. Comput. Phys. Commun. 2016, 209, 163–170. [Google Scholar] [CrossRef]
- Wang, Y. Frequencies of the ricker wavelet. Geophysics 2015, 80, A31–A37. [Google Scholar] [CrossRef]
- Qin, C.; Zhou, J.; Xu, Y.; Duan, C.; Cui, C.; Zhang, H. Using thermal infrared imaging to estimate soil moisture dynamics. J. Irrig. Drain. 2024, 43, 34–42. [Google Scholar]
- He, H.; Qi, Y.; Lü, J.; Peng, P.; Ziran, Y.; Tang, Z.; Cui, K.; Zhang, K. Development and revision of the chinese soil texture classification system. J. Agric. Resour. Environ. 2023, 40, 501. [Google Scholar]
- Peplinski, N.R.; Ulaby, F.T.; Dobson, M.C. Dielectric properties of soils in the 0.3–1.3-GHz range. IEEE Trans. Geosci. Remote. Sens. 1995, 33, 803–807. [Google Scholar] [CrossRef]
- Giannakis, I. Realistic Numerical Modelling of Ground Penetrating Radar for Landmine Detection. Ph.D. Dissertation, The University of Edinburgh, Edinburgh, UK, 2016. [Google Scholar]
- Benedetto, A.; Tosti, F.; Ciampoli, L.B.; D’amico, F. An overview of ground-penetrating radar signal processing techniques for road inspections. Signal Process. 2017, 132, 201–209. [Google Scholar] [CrossRef]
- Bai, X.; Luo, X.; Guo, S.; Wang, L.; Chen, H.; Mi, H.; Gao, Y. A novel gain control method based on extremum envelope for high speed array GPR. In Proceedings of the 2020 IEEE 92nd Vehicular Technology Conference (VTC2020-Fall), Virtual, 18 November–16 December 2020. [Google Scholar]
- Rukundo, O.; Cao, H. Nearest neighbor value interpolation. Int. J. Adv. Comput. Sci. Appl. 2012, 3, 25–30. [Google Scholar]
- Chicco, D.; Warrens, M.J.; Jurman, G. The coefficient of determination r-squared is more informative than smape, mae, mape, mse and rmse in regression analysis evaluation. PeerJ Comput. Sci. 2021, 7, e623. [Google Scholar] [CrossRef]
- Setiadi, D.R.I.M. PSNR vs SSIM: Imperceptibility quality assessment for image steganography. Multimed. Tools Appl. 2021, 80, 8423–8444. [Google Scholar] [CrossRef]
- Zhou, P.; Xie, X.; Lin, Z.; Yan, S. Towards understanding convergence and generalization of AdamW. IEEE Trans. Pattern Anal. Mach. Intell. 2024, 9, 6486–6493. [Google Scholar] [CrossRef]
- Johnson, O.V.; Xinying, C.; Khaw, K.W.; Lee, M.H. ps-CALR: Periodic-Shift Cosine Annealing Learning Rate for Deep Neural Networks. IEEE Access 2023, 11, 139171–139186. [Google Scholar] [CrossRef]
- Bergstrom, A.C.; Conran, D.; Messinger, D.W. Gaussian blur and relative edge response. arXiv 2023, arXiv:2301.00856. [Google Scholar]
- Topp, G.C.; Davis, J.L.; Annan, A.P. Electromagnetic determination of soil water content: Measurements in coaxial transmission lines. Water Resour. Res. 1980, 16, 574–582. [Google Scholar] [CrossRef]
- Rajulu, K.V.; Mohanty, B.N. Dielectric and conductivity properties of some wood composites. Int. J. Eng. Technol. 2016, 8, 51–60. [Google Scholar]
- Iman, M.; Arabnia, H.R.; Rasheed, K. A review of deep transfer learning and recent advancements. Technologies 2023, 11, 40. [Google Scholar] [CrossRef]
Model | MSE () ↑ | MAE () ↑ | SSIM ↓ | Prediction Time(s) | Number of Parameters () |
---|---|---|---|---|---|
PyViTENet | 5.18 | 2.2760 | 0.9453795 | 1.14 | 18.9 |
EDMFEBs | 5.74 | 2.3958 | 0.9364224 | 0.88 | 7.9 |
U-Net | 9.99 | 3.1612 | 0.9158119 | 0.83 | 1.9 |
TransUNet | 15.98 | 3.9975 | 0.8852904 | 1.12 | 66.8 |
GPRInvNet | 21.93 | 4.6829 | 0.8367578 | 0.85 | 1.4 |
DMRF-UNet | 47.16 | 6.8672 | 0.5886880 | 0.91 | 7.7 |
Model | MSE (10−4) | MAE (10−2) | SSIM |
---|---|---|---|
PyConvFEBs | 5.85 | 2.4195 | 0.9408528 |
PyConvFEBs + ViTFEB | 5.54 | 2.3537 | 0.9442102 |
PyConvFEBs + Edge Task ( = 1, = 1) | 8.74 | 2.9563 | 0.9120258 |
PyConvFEBs + Edge Task ( = 1, = 0.01) | 5.70 | 2.4021 | 0.9422152 |
PyViTENet ( = 1, = 1) | 8.36 | 2.8714 | 0.9361667 |
PyViTENet ( = 1, = 0.01) | 5.18 | 2.2760 | 0.9453795 |
Model | MSE (10−4) ↑ | MAE (10−2) ↑ | SSIM ↓ |
---|---|---|---|
PyViTENet | 3.31 | 1.1618 | 0.9715505 |
EDMFEBs | 3.45 | 1.6213 | 0.9715072 |
TransUNet | 93.40 | 9.64 | 0.7297934 |
U-Net | 297.42 | 15.1606 | 0.6747670 |
GPRInvNet | 1260.75 | 30.2256 | 0.5292989 |
DMRF-UNet | 2416.36 | 48.4334 | 0.4273940 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Li, X.; Cheng, X.; Zhao, Y.; Xiang, B.; Zhang, T. Deep Learning-Based Ground-Penetrating Radar Inversion for Tree Roots in Heterogeneous Soil. Sensors 2025, 25, 947. https://doi.org/10.3390/s25030947
Li X, Cheng X, Zhao Y, Xiang B, Zhang T. Deep Learning-Based Ground-Penetrating Radar Inversion for Tree Roots in Heterogeneous Soil. Sensors. 2025; 25(3):947. https://doi.org/10.3390/s25030947
Chicago/Turabian StyleLi, Xibei, Xi Cheng, Yunjie Zhao, Binbin Xiang, and Taihong Zhang. 2025. "Deep Learning-Based Ground-Penetrating Radar Inversion for Tree Roots in Heterogeneous Soil" Sensors 25, no. 3: 947. https://doi.org/10.3390/s25030947
APA StyleLi, X., Cheng, X., Zhao, Y., Xiang, B., & Zhang, T. (2025). Deep Learning-Based Ground-Penetrating Radar Inversion for Tree Roots in Heterogeneous Soil. Sensors, 25(3), 947. https://doi.org/10.3390/s25030947