面向探地雷达常见地下目标的GDS-YOLOv8n检测方法

王惠琴 杨发东 何永强 刘宾灿 刘鑫

王惠琴, 杨发东, 何永强, 等. 面向探地雷达常见地下目标的GDS-YOLOv8n检测方法[J]. 雷达学报(中英文), 待出版. doi: 10.12000/JR24160
引用本文: 王惠琴, 杨发东, 何永强, 等. 面向探地雷达常见地下目标的GDS-YOLOv8n检测方法[J]. 雷达学报(中英文), 待出版. doi: 10.12000/JR24160
WANG Huiqin, YANG Fadong, HE Yongqiang, et al. Detection of common underground targets in ground penetrating radar images using the GDS-YOLOv8n model[J]. Journal of Radars, in press. doi: 10.12000/JR24160
Citation: WANG Huiqin, YANG Fadong, HE Yongqiang, et al. Detection of common underground targets in ground penetrating radar images using the GDS-YOLOv8n model[J]. Journal of Radars, in press. doi: 10.12000/JR24160

面向探地雷达常见地下目标的GDS-YOLOv8n检测方法

DOI: 10.12000/JR24160
基金项目: 甘肃省重点研发计划 (23YFFA0060),甘肃省优秀研究生“创新之星”项目
详细信息
    作者简介:

    王惠琴,博士,教授,博士生导师,主要研究方向为雷达信号检测与处理、无线光通信理论与技术等

    杨发东,硕士生,主要研究方向为雷达信号处理和深度学习

    何永强,博士,教授,硕士生导师,主要研究方向为探地雷达信号智能处理、地质灾害分析与防治技术等

    刘宾灿,高级工程师,主要研究方向为隔震建筑的监测与性能评估、建筑节能及碳排放核算等

    刘 鑫,硕士生,主要研究方向为探地雷达信号的去噪和反演

    通讯作者:

    王惠琴 whq1222@lut.edu.cn

    杨发东 1948374001@qq.com

  • 责任主编:雷文太 Corresponding Editor: LEI Wentai
  • 中图分类号: TN957; P631

Detection of Common Underground Targets in Ground Penetrating Radar Images Using the GDS-YOLOv8n Model

Funds: Key Research and Development Program of Gansu Province (23YFFA0060), “Innovation Star” Project of Outstanding Graduate Students in Gansu Province
More Information
  • 摘要: 针对当前探地雷达(GPR)图像检测中存在准确率低、误检和漏检等问题,该文提出了一种GPR常见地下目标检测模型GDS-YOLOv8n。该模型首先使用DRRB特征提取模块替换YOLOv8n模型中的部分C2f模块,旨在增强模型对多尺度特征的提取能力。其次使用SPD-Conv下采样模块替换像素为320×320及以下特征图所对应的Conv模块,有效克服分辨率受限以及存在小目标的GPR图像在下采样过程中的信息损失问题;同时利用辅助训练模块,在不增加检测阶段模型复杂度的前提下提升GPR图像的检测性能。最后,引入Inner-SIoU损失函数,在添加新约束条件的基础上,通过比例因子生成适合于当前GPR图像的辅助边界框,以提高预测框的准确性。实验结果表明,GDS-YOLOv8n模型对金属管、PVC管和电缆线等6类常见地下目标在实测GPR图像数据集上的P, R和mAP50分别为97.1%, 96.2%和96.9%,较YOLOv8n模型分别提高了4.0%, 6.1%和4.1%,尤其对PVC管和电缆线目标的检测效果提升更明显。与YOLOv5n, YOLOv7-tiny和SSD等模型相比,其mAP50分别提高了7.20%, 5.70%和14.48%。此外,将GDS-YOLOv8n模型部署到NVIDIA Jetson Orin NX嵌入式设备上,并使用TensorRT进行优化。经FP16量化后,模型的检测速度由22.0 FPS提高到40.6 FPS,能够满足移动场景下GPR地下目标实时探测任务的需求。

     

  • 图  1  试验数据采集部分场景

    Figure  1.  Partial scenarios for experimental data collection

    图  2  实测B-scan图片

    Figure  2.  Measured B-scan image

    图  3  GDS-YOLOv8n网络结构图

    Figure  3.  GDS-YOLOv8n network structure diagram

    图  4  DRRB模块

    Figure  4.  DRRB module

    图  5  DRB模块

    Figure  5.  DRB module

    图  6  结构重参数化过程

    Figure  6.  Structural reparameterization process

    图  7  有效感受野

    Figure  7.  Effective receptive field

    图  8  SPD-Conv模块

    Figure  8.  SPD-Conv module

    图  9  带辅助训练头的训练方式

    Figure  9.  Training method of auxiliary head

    图  10  训练过程曲线

    Figure  10.  Training process curve

    图  11  GPR B-scan图像检测结果对比

    Figure  11.  Comparison of GPR B-scan image detection results

    图  12  测试集上YOLOv8n模型的精确度-召回率曲线

    Figure  12.  P-R curve of YOLOv8n model on the test dataset

    图  13  测试集上GSD-YOLOv8n模型的精确度-召回率曲线

    Figure  13.  P-R curve of GSD-YOLOv8n model on the test dataset

    图  14  不同模型在测试集上的精确度-召回率曲线

    Figure  14.  P-R curves of different models on the test set

    图  15  模型综合性能比较

    Figure  15.  Comprehensive performance comparison of models

    图  16  模型部署流程

    Figure  16.  Model deployment process

    表  1  DRB模块参数

    Table  1.   DRB module parameter

    Size k d
    5 3, 3 1, 2
    7 5, 5, 3 1, 2, 3
    9 5, 5, 3, 3 1, 2, 3, 4
    11 5, 5, 3, 3, 3 1, 2, 3, 4, 5
    13 5, 7, 3, 3, 3 1, 2, 3, 4, 5
    下载: 导出CSV

    表  2  DRRB模块性能对比

    Table  2.   DRRB module performance comparison

    20×20 40×40 Params (Million) mAP50 (%)
    (5, 7) (7, 9) 2.690 93.9
    (5, 7) (9, 11) 2.699 94.0
    (7, 9) (9, 11) 2.706 95.3
    (7, 9) (11, 13) 2.717 94.8
    (9, 11) (11, 13) 2.727 94.1
    注:(x, y)表示使用尺寸为xy的DRB模块。
    下载: 导出CSV

    表  3  模型训练参数设置

    Table  3.   Model training parameter settings

    参数名称参数设置
    迭代次数300
    批大小16
    早停机制50
    初始学习率0.01
    最终学习率0.0001
    动量0.937
    优化器权重衰减0.0005
    优化器auto
    注:优化器参数设置auto时,自动选择AdamW优化器。
    下载: 导出CSV

    表  4  算法改进前后P, R值对比(%)

    Table  4.   Comparison of P and R values before and after algorithm improvement (%)

    模型 金属管 PVC 电缆线 含水塑料瓶 注浆不密实
    加固体空洞
    管网 总计
    P R P R P R P R P R P R P R
    YOLOv8n 95.5 82.1 93.4 83.2 100 85.1 93.2 95.3 97.4 96.5 79.5 98.4 93.1 90.1
    GDS-YOLOv8n 96.3 87.2 97.5 98.3 100 100 100 100 97.7 93 91.2 98.4 97.1 96.2
    注:加粗数字表示最优值。
    下载: 导出CSV

    表  5  消融实验结果(%)

    Table  5.   Results of ablation experiments (%)

    实验 模型 P R mAP50 mAP50:95
    1 YOLOv8n 93.1 90.1 92.8 60.7
    2 YOLOv8n+A 95.1 93.2 95.3 63.4
    3 YOLOv8n+B 94.6 92.6 94.4 61.3
    4 YOLOv8n+C 96.5 91.3 94.7 61.9
    5 YOLOv8n+L (ratio=1.29) 95.2 92.0 94.3 61.8
    6 YOLOv8n+A+B 96.4 94.5 96.0 64.0
    7 YOLOv8n+A+B+C 96.8 94.4 96.5 64.2
    8 YOLOv8n+A+B+C+L (ratio=1.28) 97.1 96.2 96.9 64.3
    注:加粗数字表示最优值,ratio是Inner-SIoU中的比例因子。
    下载: 导出CSV

    表  6  对比实验结果

    Table  6.   Results of comparison experiments

    模型 Params (Million) GFLOPs Model size (MB) P (%) R (%) mAP50 (%) mAP50:90 (%) FPS (bs=1)
    SSD* 24.28 176.23 93.10 84.03 81.86 82.42 48.30 35.20
    YOLOv5n 1.77 4.20 3.68 89.30 86.00 89.70 61.10 129.90
    YOLOv5s 7.03 15.80 13.70 90.70 90.40 92.30 62.80 123.50
    YOLOv7-tiny 6.03 13.10 11.70 91.90 92.60 91.20 57.60 80.00
    YOLOv8n 3.01 8.10 5.94 93.10 90.10 92.80 60.70 109.90
    YOLOv8s 11.13 28.40 21.40 94.70 90.70 93.70 62.00 103.10
    YOLOv8m 25.80 78.70 49.50 94.10 96.30 96.70 63.30 55.60
    GDS-YOLOv8n 4.43 11.20 10.30 97.10 96.20 96.90 64.30 97.20
    注:加粗数字表示最优值;bs=1表示 batch size取1时的值;$* $表示输入图片尺寸是512×512。
    下载: 导出CSV

    表  7  GDS-YOLOv8n模型部署格式性能对比

    Table  7.   Performance comparison of GDS-YOLOv8n model deployment formats

    模型 Model size
    (MB)
    mAP50
    (%)
    FPS
    (bs=1)
    GDS-YOLOv8n.pt 10.3 96.9 22.0
    GDS-YOLOv8n.onnx 17.1 96.9 6.5
    GDS-YOLOv8n.engine (FP32) 19.1 96.9 31.4
    GDS-YOLOv8n.engine (FP16) 10.7 96.9 40.6
    GDS-YOLOv8n.engine (INT8) 5.6 84.1 46.0
    注:FP32, FP16和INT8分别表示以32位、16位浮点数和8位整数量化模型。
    下载: 导出CSV
  • [1] LIU Wenchao, LUO Rong, XIAO Manzhe, et al. Intelligent detection of hidden distresses in asphalt pavement based on GPR and deep learning algorithm[J]. Construction and Building Materials, 2024, 416: 135089. doi: 10.1016/j.conbuildmat.2024.135089.
    [2] ZHU Jiasong, ZHAO Dingyi, and LUO Xianghuan. Evaluating the optimised YOLO-based defect detection method for subsurface diagnosis with ground penetrating radar[J]. Road Materials and Pavement Design, 2024, 25(1): 186–203. doi: 10.1080/14680629.2023.2199880.
    [3] RICHARDS E, STUEFER S, RANGEL R C, et al. An evaluation of GPR monitoring methods on varying river ice conditions: A case study in Alaska[J]. Cold Regions Science and Technology, 2023, 210: 103819. doi: 10.1016/j.coldregions.2023.103819.
    [4] WANG Xiaofen, WANG Peng, ZHANG Xiaotong, et al. Target electromagnetic detection method in underground environment: A review[J]. IEEE Sensors Journal, 2022, 22(14): 13835–13852. doi: 10.1109/JSEN.2022.3175502.
    [5] KAUR P, DANA K J, ROMERO F A, et al. Automated GPR rebar analysis for robotic bridge deck evaluation[J]. IEEE Transactions on Cybernetics, 2016, 46(10): 2265–2276. doi: 10.1109/TCYB.2015.2474747.
    [6] SARKAR M, SR N, HEMANI M, et al. Parameter efficient local implicit image function network for face segmentation[C]. The IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023: 20970–20980. doi: 10.1109/CVPR52729.2023.02009.
    [7] TONG Zheng, GAO Jie, and YUAN Dongdong. Advances of deep learning applications in ground-penetrating radar: A survey[J]. Construction and Building Materials, 2020, 258: 120371. doi: 10.1016/j.conbuildmat.2020.120371.
    [8] REN Shaoqing, HE Kaiming, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137–1149. doi: 10.1109/TPAMI.2016.2577031.
    [9] LIU Wei, ANGUELOV D, ERHAN D, et al. SSD: Single shot multibox detector[C]. The 14th European Conference on Computer Vision, Amsterdam, The Netherlands, 2016: 21–37. doi: 10.1007/978-3-319-46448-0_2.
    [10] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real-time object detection[C]. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016: 779–788. doi: 10.1109/CVPR.2016.91.
    [11] 杨必胜, 宗泽亮, 陈驰, 等. 车载探地雷达地下目标实时探测法[J]. 测绘学报, 2020, 49(7): 874–882. doi: 10.11947/j.AGCS.2020.20190293.

    YANG Bisheng, ZONG Zeliang, CHEN Chi, et al. Real time approach for underground objects detection from vehicle-borne ground penetrating radar[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(7): 874–882. doi: 10.11947/j.AGCS.2020.20190293.
    [12] XIONG Xuetang and TAN Yiqiu. Deep learning-based detection of tie bars in concrete pavement using ground penetrating radar[J]. International Journal of Pavement Engineering, 2023, 24(2): 2155648. doi: 10.1080/10298436.2022.2155648.
    [13] HU Haobang, FANG Hongyuan, WANG Niannian, et al. Defects identification and location of underground space for ground penetrating radar based on deep learning[J]. Tunnelling and Underground Space Technology, 2023, 140: 105278. doi: 10.1016/j.tust.2023.105278.
    [14] 王惠琴, 罗佳, 何永强, 等. 改进YOLOv5的探地雷达常见地下管线识别[J]. 地球物理学报, 2024, 67(9): 3588–3604. doi: 10.6038/cjg2023R0431.

    WANG Huiqin, LUO Jia, HE Yongqiang, et al. Identification of common underground pipelines by ground penetrating radar based on improved YOLOv5[J]. Chinese Journal of Geophysics, 2024, 67(9): 3588–3604. doi: 10.6038/cjg2023R0431.
    [15] 刘震, 顾兴宇, 李骏, 等. 探地雷达数值模拟与道路裂缝图像检测的深度学习增强方法[J]. 地球物理学报, 2024, 67(6): 2455–2471. doi: 10.6038/cjg2023R0090.

    LIU Zhen, GU Xingyu, LI Jun, et al. Deep learning-enhanced numerical simulation of ground penetrating radar and image detection of road cracks[J]. Chinese Journal of Geophysics, 2024, 67(6): 2455–2471. doi: 10.6038/cjg2023R0090.
    [16] DING Xiaohan, ZHANG Xiangyu, HAN Jungong, et al. Scaling up your kernels to 31×31: Revisiting large kernel design in CNNs[C]. The IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, USA, 2022: 11953–11965. doi: 10.1109/CVPR52688.2022.01166.
    [17] WEI Haoran, LIU Xu, Xu Shouchun, et al. DWRSeg: Rethinking efficient acquisition of multi-scale contextual information for real-time semantic segmentation[J]. arXiv: 2212.01173, 2022. doi: 10.48550/arXiv.2212.01173.
    [18] DING Xiaohan, ZHANG Yiyuan, GE Yixiao, et al. UniRepLKNet: A universal perception large-kernel convnet for audio, video, point cloud, time-series and image recognition[C]. The IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, USA, 2024: 5513–5524. doi: 10.1109/CVPR52733.2024.00527.
    [19] SUNKARA R and LUO Tie. No more strided convolutions or pooling: A new CNN building block for low-resolution images and small objects[C]. Joint European Conference on Machine Learning and Knowledge Discovery in Databases, Grenoble, France, 2023: 443–459. doi: 10.1007/978-3-031-26409-2_27.
    [20] WANG C Y, BOCHKOVSKIY A, and LIAO H Y M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]. The IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, 2023: 7464–7475. doi: 10.1109/CVPR52729.2023.00721.
    [21] ZHANG Hao, XU Cong, and ZHANG Shuaijie. Inner-IoU: More effective intersection over union loss with auxiliary bounding box[J]. arXiv: 2311.02877, 2023. doi: 10.48550/arXiv.2311.02877.
  • 加载中
图(16) / 表(7)
计量
  • 文章访问数:  204
  • HTML全文浏览量:  76
  • PDF下载量:  28
  • 被引次数: 0
出版历程
  • 收稿日期:  2024-08-09
  • 修回日期:  2024-10-14
  • 网络出版日期:  2024-10-24

目录

    /

    返回文章
    返回