面向路面附着估计的路面图像识别
作者:
作者单位:

江西理工大学 电气工程与自动化学院,江西 赣州 341000

作者简介:

黄开启(1969—),男,教授,主要从事新能源汽车与机器人控制技术研究,(E-mail) kaiqi.huang@163.com。

中图分类号:

TP391.4

基金项目:

国家自然科学基金资助项目(61963018)。


Road image recognition for road adhesion estimation
Author:
Affiliation:

School of Electrical Engineering and Automation, Jiangxi University of Science and Technology, Ganzhou, Jiangxi 341000, P. R. China

Fund Project:

Supported by National Natural Science Foundation of China (61963018).

  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献 [21]
  • |
  • 相似文献 [20]
  • |
  • 引证文献
  • | |
  • 文章评论
    摘要:

    为提升智能辅助驾驶系统对路面附着系数估计的准确性与实时性,研究了一种基于视觉信息的路面识别深度学习算法,实现路面附着系数的预估计。设计压缩卷积机制以降低网络运算参数,采用特征图全局平均替换全连接层以提升网络的拟合性能,并构建路面识别深度卷积神经网络DW-VGG。利用自建路面图像数据集对网络进行训练,测试结果表明,基于提出的多层知识蒸馏技术的DW-VGG网络识别精度较高,分类性能评估指标F1得分为96.57%,并有效降低了网络的运算和内存成本,识别单张图像只需32.06 ms,预测模型只有5.63 M。

    Abstract:

    To enhance the accuracy and real-time performance of the intelligent assisted driving system in estimating the road adhesion coefficient, a deep learning algorithm based on visual information was developed for road recognition. The algorithm aims to achieve a pre-estimation of the road adhesion coefficient. A compression convolution mechanism was designed to reduce the network’s operation parameters. Additionally, the fully connection layer was replaced by the global average of the feature map to enhance the network’s fitting performance. Furthermore, a pavement recognition depth convolutional neural network called DW-VGG was constructed. The network was trained using a self-built pavement image dataset. The test results demonstrate that the DW-VGG network, utilizing the proposed multi-layer knowledge distillation algorithm, achieves a high recognition accuracy, with a classification performance evaluation index (F1 score) of 96.57%. Moreover, it effectively reduces the network’s time and space costs, as it only takes 32.06 ms to identify a single image, and the prediction model size is merely 5.63 M.

    参考文献
    [1] Rajendran S, Spurgeon S K, Tsampardoukas G, et al. Estimation of road frictional force and wheel slip for effective antilock braking system (ABS) control[J]. International Journal of Robust and Nonlinear Control, 2019, 29(3): 736-765.
    [2] Han K, Lee B, Choi S B. Development of an antilock brake system for electric vehicles without wheel slip and road friction information[J]. IEEE Transactions on Vehicular Technology, 2019, 68(6): 5506-5517.
    [3] Khaleghian S, Emami A, Taheri S. A technical survey on tire-road friction estimation[J]. Friction, 2017, 5(2): 123-146.
    [4] 余卓平, 曾德全, 熊璐, 等. 基于激光雷达的无人车路面附着系数估计[J]. 华中科技大学学报(自然科学版), 2019, 47(7): 124-127.Yu Z P, Zeng D Q, Xiong L, et al. Road adhesion coefficient estimation for unmanned vehicle based on lidar[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2019, 47(7): 124-127.(in Chinese)
    [5] Fényes D, Németh B, Gáspár P, et al. Road surface estimation based LPV control design for autonomous vehicles[J]. IFAC-PapersOnLine, 2019, 52(28): 120-125.
    [6] Paul D, Velenis E, Cao D, et al. Optimal μ-estimation-based regenerative braking strategy for an AWD HEV[J]. IEEE Transactions on Transportation Electrification, 2017,3(1): 249-258.
    [7] Ping X Y, Cheng S, Yue W, et al. Adaptive estimations of tyre-road friction coefficient and body’s sideslip angle based on strong tracking and interactive multiple model theories[J]. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2020, 234(14): 3224-3238.
    [8] Senatore A, Sharifzadeh M. Estimation of tyre-road friction during ABS braking for snow and ice conditions [C]//6° Workshop dell’Associazione Italiana. di Tribologia AIT Torino, Italy, 2018.
    [9] Sentouh C, Nguyen A T, Benloucif M A, et al. Driver-automation cooperation oriented approach for shared control of lane keeping assist systems[J]. IEEE Transactions on Control Systems Technology, 2019, 27(5): 1962-1978.
    [10] 冯加明, 储茂祥, 杨永辉, 等. 改进YOLOv3算法的车辆信息检测[J]. 重庆大学学报, 2021, 44(12): 71-79.Feng J M, Chu M X, Yang Y H, et al. Vehicle information detection based on improved YOLOv3 algorithm[J]. Journal of Chongqing University, 2021, 44(12): 71-79.(in Chinese)
    [11] Wang Y F, Nguyen B M, Fujimoto H, et al. Multirate estimation and control of body slip angle for electric vehicles based on onboard vision system[J]. IEEE Transactions on Industrial Electronics, 2014, 61(2): 1133-1143.
    [12] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[EB/OL]. 2014: arXiv: 1409.1556. https://arxiv.org/abs/1409.1556.
    [13] Geiger A, Lenz P, Stiller C, et al. Vision meets robotics: the KITTI dataset[J]. The International Journal of Robotics Research, 2013, 32(11): 1231-1237.
    [14] Yu F, Chen H F, Wang X, et al. BDD100K: a diverse driving dataset for heterogeneous multitask learning[EB/OL]. 2018: arXiv: 1805.04687. https://arxiv.org/abs/1805.04687.
    [15] Maddern W, Pascoe G, Linegar C, et al. 1 year, 1000 km: the Oxford RobotCar dataset[J]. The International Journal of Robotics Research, 2017, 36(1): 3-15.
    [16] Szegedy C, Vanhoucke V, Ioffe S, et al. Rethinking the inception architecture for computer vision[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). IEEE, 2016: 2818-2826.
    [17] He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2016: 770-778.
    [18] Sandler M, Howard A, Zhu M L, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 2018: 4510-4520.
    [19] Iandola F N, Han S, Moskewicz M W, et al. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size[EB/OL]. 2016: arXiv: 1602.07360. https://arxiv.org/abs/1602.07360.
    [20] Ma N N, Zhang X Y, Zheng H T, et al. ShuffleNet V2: practical guidelines for efficient CNN architecture design[C]//European Conference on Computer Vision. Cham: Springer, 2018: 122-138.
    [21] Tan M, Le Q V. EfficientNet: Rethinking model scaling for convolutional neural networks[C]//2019 International Conference on Machine Learning (ICML). IEEE, 2019.
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

黄开启,黄茂云,刘小荣.面向路面附着估计的路面图像识别[J].重庆大学学报,2023,46(7):97-106.

复制
分享
文章指标
  • 点击次数:383
  • 下载次数: 683
  • HTML阅读次数: 192
  • 引用次数: 0
历史
  • 收稿日期:2021-06-08
  • 在线发布日期: 2023-08-02
文章二维码