|
|
基于改进TransUNet的沥青路面病害检测研究
|
Abstract:
本文研究了基于图像分割算法的沥青路面病害检测问题，提出了一种基于TransUNet的分割网络，用于沥青路面病害的检测和识别。该网络在TransUNet的骨干网络中引入了ECA注意力机制，以增强特征的表达能力和自适应性。本文使用Dice评价系数作为评价指标，对路面裂缝和坑洞进行分割，并在公开的路面病害数据集上进行了实验。实验结果表明，所提出的TransUNet-RD模型mIoU和Dice等方面优于原始的TransUNet模型和UNet模型，证明了其在沥青路面病害检测方面的有效性和优越性。
This paper studies the problem of asphalt pavement disease detection based on image segmentation algorithms, and proposes a segmentation network based on TransUNet for asphalt pavement disease detection and recognition. The network introduces the ECA attention mechanism in the back-bone network of TransUNet, to enhance the feature expression ability and adaptability. This paper uses the Dice coefficient as the evaluation index, segments the pavement cracks and potholes, and conducts experiments on the public pavement disease dataset. The experimental results show that the proposed TransUNet-RD model is superior to the original TransUNet model and UNet model in terms of mIoU and Dice, proving its effectiveness and superiority in asphalt pavement distress detection.
| [1] | Girshick, R., Donahue, J., Darrell, T., et al. (2014) Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation. 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, 23-28 June 2014, 580-587. https://doi.org/10.1109/CVPR.2014.81 |
| [2] | Girshick, R. (2015) Fast R-CNN. 2015 IEEE International Conference on Computer Vision (ICCV), Santiago, 07-13 December 2015, 1440-1448. https://doi.org/10.1109/ICCV.2015.169 |
| [3] | Ren, S., He, K., Girshick, R., et al. (2015) Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Advances in Neural Information Processing Systems, 28, 1137-1149. |
| [4] | Alzraiee, H., Leal Ruiz, A. and Sprotte, R. (2021) Detecting of Pavement Marking Defects Using Faster R-CNN. Journal of Performance of Constructed Facilities, 35, Article 04021035.
https://doi.org/10.1061/(ASCE)CF.1943-5509.0001606 |
| [5] | Duan, K., Bai, S., Xie, L., et al. (2019) Centernet: Keypoint Triplets for Object Detection. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, 27 October-02 November 2019, 6569-6578.
https://doi.org/10.1109/ICCV.2019.00667 |
| [6] | Redmon, J., Divvala, S., Girshick, R., et al. (2016) You Only Look Once: Unified, Real-Time Object Detection. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 779-788.
https://doi.org/10.1109/CVPR.2016.91 |
| [7] | Redmon, J. and Farhadi, A. (2018) YOLOv3: An Incremental Improvement.
https://arxiv.org/abs/1804.02767 |
| [8] | Bochkovskiy, A., Wang, C.Y. and Liao, H.Y.M. (2020) YOLOv4: Optimal Speed and Accuracy of Object Detection.
https://arxiv.org/abs/2004.10934 |
| [9] | Xue, Y., et al. (2020) YOLONG: A YOLO-Based Approach for Road Crack Detection. IEEE Access, 8, 133235-133245. |
| [10] | Rahman, S.M.M., et al. (2021) Deep Learning Based Asphalt Pavement Disease Detection Using YOLOv4. IET Intelligent Transport Systems, 15, 229-236. |
