Recognition of Road Subsurface Cracks in Pavement Based on 3D Ground Penetrating Radar Multi-view Fusion

NAVIGATE

Table of Contents

STATISTICS

Viewed46

Downloads66

Recognition of Road Subsurface Cracks in Pavement Based on 3D Ground Penetrating Radar Multi-view Fusion

PDF下载 (66)

[1]ZHONG Yanhui,SHAO Yanhao,ZHANG Ruifeng,et al.Recognition of Road Subsurface Cracks in Pavement Based on 3D Ground Penetrating Radar Multi-view Fusion[J].Journal of Zhengzhou University (Engineering Science),2027,48(XX):1-8.[doi:10.13705/j. issn.1671-6833.2027.01.005]

Copy

Journal of Zhengzhou University (Engineering Science)[ISSN 1671-6833/CN 41-1339/T] Volume: 48 Number of periods: 2027 XX Page number: 1-8 Column: Public date: 2027-12-10

Title:: Recognition of Road Subsurface Cracks in Pavement Based on 3D Ground Penetrating Radar Multi-view Fusion

Author(s):: ZHONG Yanhui¹, SHAO Yanhao¹, ZHANG Ruifeng², LI Qingbo², WANG Xiang¹, ZHANG Bei¹; 1. School of Water Conservancy and Transportation, Zhengzhou University, Zhengzhou 450001, China; 2. Kaifeng Branch, Henan Communications Investment Group Co., Ltd, Kaifeng 475000, China

Keywords:: ground penetrating radar; pavement crack; multi-view fusion; attention mechanism; deep learning

CLC:: TP391.41 U416.217

DOI:: 10.13705/j. issn.1671-6833.2027.01.005

Abstract:: To address the problem that existing methods rely predominantly on single-view radar images while ignoring the complementary multi-view information of three-dimensional ground penetrating radar, a road subsurface crack identification method based on multi-view fusion and deep learning was developed. A three-dimensional forward model of asphalt pavement was established using gprMax to obtain B-scan, C-scan, and D-scan radar profiles from three orthogonal directions, and a dataset comprising 3 437 triple-view samples was constructed by combining forward simulation with field measurement data. DenseNet121 and ResNet50 were adopted as backbone networks, with a Squeeze-and-Excitation channel attention mechanism introduced, a triple-branch parallel structure constructed, and an inter-view attention fusion module designed for adaptive weighted fusion of multi-view features. Experimental results showed that the DenseNet121-TB model achieved accuracy, precision, recall, and F1-score of 98.60%, 98.71%, 98.60%, and 98.65%, respectively, with accuracy 4.19 percentage points higher than that of the best single-view model. Ablation experiments verified the positive contribution of each module to recognition performance. Field core drilling validation demonstrated good correspondence between the three-view features and the drilling results, indicating that the proposed method has practical engineering applicability.

References:: [1]. 沙爱民. 半刚性基层的材料特性[J]. 中国公路学报, 2008, 21(1): 1-5.
[2]. SHA A M. Material characteristics of semi-rigid base[J]. China Journal of Highway and Transport, 2008, 21(1): 1-5.
[3]. 李法滨,杨莉,王圣棋,等. 高速公路沥青路面裂缝分类的集合表达及其维度排序方法[J]. 公路, 2025, 70(5): 7-12.
[4]. LI F B , YANG L, WANG S Q, et al. Set expression and dimension ranking method for crack classification of expressway asphalt pavement[J]. Highway, 2025, 70(5): 7-12.
[5]. LIU H, CHEN J, ZHANG X, et al. Collaborative imaging of subsurface cavities using ground-pipeline penetrating radar[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21: 1-5.
[6]. 方宏远,董智峰,薛冰寒,等. 高聚物注浆修复的面板堤坝探地雷达波场特征分析[J]. 郑州大学学报(工学版), 2024, 45(3): 1-6, 13.
[7]. FANG H Y, DONG Z F, XUE B H, et al. Analysis of ground penetrating radar wave field characteristics of damface disengaging repaired by polymer grouting[J]. Journal of Zhengzhou University (Engineering Science), 2024, 45(3): 1-6, 13.
[8]. LIU H, YUE Y, LIU C, et al. Automatic recognition and localization of underground pipelines in GPR B-scans using a deep learning model[J]. Tunnelling and Underground Space Technology, 2023, 134: 104861.
[9]. DIAMANTI N, REDMAN D. Field observations and numerical models of GPR response from vertical pavement cracks[J]. Journal of Applied Geophysics, 2012, 81: 106-116.
[10]. 卢成明,秦臻,朱海龙,等. 探地雷达检测公路结构层隐含裂缝实用方法研究[J]. 地球物理学报, 2007, 50(5): 1558-1568.
[11]. LU C M, QIN Z, ZHU H L, et al. Practical method for detecting hidden cracks in highway structural layers using ground penetrating radar[J]. Chinese Journal of Geophysics, 2007, 50(5): 1558-1568.
[12]. TONG Z, GAO J, ZHANG H. Innovative method for recognizing subgrade defects based on a convolutional neural network[J]. Construction and Building Materials, 2018, 169: 69-82.
[13]. LIU Z, YEOH J K W, GU X, et al. Automatic pixel-level detection of vertical cracks in asphalt pavement based on GPR investigation and improved mask R-CNN[J]. Automation in Construction, 2023, 146: 104689.
[14]. QUE Y, DAI Y, JI X, et al. Automatic classification of asphalt pavement cracks using a novel integrated generative adversarial network and improved VGG model[J]. Engineering Structures, 2023, 277: 115406.
[15]. KIM N, KIM S, AN Y K, et al. A novel 3D GPR image arrangement for deep learning-based underground object classification[J]. International Journal of Pavement Engineering, 2021, 22(6): 740-751.
[16]. KIM N, KIM S, AN Y K, et al. Triplanar imaging of 3-D GPR data for deep-learning-based underground object detection[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, 12(11): 4446-4456.
[17]. LI H, LI N, WU R, et al. GPR-RCNN: an algorithm of subsurface defect detection for airport runway based on GPR[J]. IEEE Robotics and Automation Letters, 2021, 6(2): 3001-3008.
[18]. LI N N, WU R, LI H F, et al. MV-GPRNet: multi-view subsurface defect detection network for airport runway inspection based on GPR[J]. Remote Sensing, 2022, 14(18): 4472.
[19]. MOALLA M, FRIGUI H, KAREM A, et al. Application of convolutional and recurrent neural networks for buried threat detection using ground penetrating radar data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(10): 7022-7034.
[20]. 韩海航,莫佳笛,周春鹏,等. 基于注意力融合的三维探地雷达道路病害检测[J]. 计算机工程与设计, 2022, 43(9): 2669-2677. HAN H H, MO J D, ZHOU C P, et al. Road distress detection of 3D ground penetrating radar based on attention fusion[J]. Computer Engineering and Design, 2022, 43(9): 2669-2677.
[21]. ZHOU C, NING K, WANG H, et al. Multi-view fusion and distillation for subgrade distresses detection based on 3D-GPR[EB/OL]. (2023-08-09)[2025-08-10].https://arxiv.org/abs/2308.04779.
[22]. YAMAGUCHI T, MIZUTANI T, NAGAYAMA T. Mapping subsurface utility pipes by 3-D convolutional neural network and Kirchhoff migration using GPR images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(8): 6525-6536.
[23]. HUANG G, LIU Z, VAN DER MAATEN L, et al. Densely connected convolutional networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 2261-2269.
[24]. HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[25]. HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132-7141.
[26]. LIU Z, LIN Y, CAO Y, et al. Swin Transformer: hierarchical vision Transformer using shifted windows[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 9992-10002.
[27]. WOO S, DEBNATH S, HU R, et al. ConvNeXt V2: co-designing and scaling ConvNets with masked autoencoders[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 16133-16142.
[28]. WANG A, CHEN H, LIN Z, et al. RepViT: revisiting mobile CNN from ViT perspective[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2024: 15909-15920.

Similar References:

Memo

Last Update: 2026-10-30