Research on the Improved YOLOv8 Algorithm for Detecting Multi - target Road Obstacles

Jiangfeng Sun; Pengfei Feng; Yanyang  Zeng

doi:10.54097/ws98nh43

Authors

Jiangfeng Sun
Pengfei Feng
Yanyang Zeng

DOI:

https://doi.org/10.54097/ws98nh43

Keywords:

Multi-object Obstacle Detection, YOLOv8, Improved C2f Module

Abstract

Different road obstacles may have diverse appearance, size, quantity, and location characteristics, and they are often accompanied by occlusion and overlapping phenomena. These factors result in the current object detection algorithms performing poorly in terms of detection accuracy and robustness when dealing with multiple objects and small objects with indistinct features. To address this issue, a lightweight road obstacle detection algorithm based on the improved YOLOv8n is proposed. Firstly, the MFObstacle road obstacle object detection dataset is created, which expands the types and quantities of obstacles for detection. Secondly, in the Neck part of the network, the GLSA (Global-to-Local Spatial Aggregation) module is added to enhance the multi-scale spatial aggregation mechanism of the network and improve the model's sensitivity and recognition ability for objects of different scales. Then, the BiFPN (Bidirectional Feature Pyramid Network) is integrated. BiFPN can dynamically adjust the weights according to the input data, making the network structure simpler and the computational efficiency higher. Finally, MPDIoU is used. MPDIoU takes into account the overlapping or non-overlapping regions, the distance between the center points, as well as the deviations in width and height, and simplifies the calculation process. Compared with the original YOLOv8n algorithm, the mAP0.5 index of this algorithm is increased by 2.2% to reach 85.7%; the number of network parameters is reduced by 30.4% to 4.6MB; and the computational complexity is reduced by 6% to 7.6GFLOPs. This algorithm improves the detection accuracy while reducing the number of parameters and computational complexity, and can be better applied to the task of multi-object road obstacle detection.

Downloads

Download data is not yet available.

References

[1] Zhang X Y, Gao H B, Zhao J F, et al. A review of deep learning based autonomous driving technologies[J]. Journal of Tsinghua University (Natural Science Edition), 2018, 58(004): 438-444.

[2] Ran Z S, Li R and He S. Road obstacle detection algorithm based on improved YOLOv5s[J]. Electric Measurement Technology, 2024, 46(22): 177-185.

[3] GIRSHICK R.Fast R-CNN[C].Proceedings of the IEEE International conference on computer vision, Santiago, Chile, 2015:1440-1448.

[4] REN S, HE K, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[C]. IEEE Transactions on Pattern Analysis and Machine Intelligence, Montréal, Canada, 2015:91-99.

[5] HUU P N, THI Q P and QUYNH P T. Proposing lane and obstacle detection algorithm using YOLO to control selfdriving cars on advanced networks [J]. Advances in Multimedia, 2022: 3425295.

[6] CAO J, SONG C, PENG S, et al. Pedestrian detection algorithm for intelligent vehicles in complex scenarios[J]. Sensors (Switzerland), 2020, 20(13): 3646.

[7] SHIVAYOGI A B, DHARMENDRA N C M, RAMAKRISHNA A M, et al. Real-time traffic sign recognition using deep learning [J]. Pertanika Journal of Science and Technology, 2023, 31(1): 137-148.

[8] WANG Y R, YANG G P, GUO J B. Vehicle detection in surveillance videos based on YOLOv5 lightweight network [J]. Bulletin of the Polish Academy of SciencesTechnical Sciences, 2022, 70(6): 143644.

[9] HUANG P, WANG S, CHEN J, et al. Lightweight model for pavement defect detection based on improved YOLOv7 [J]. Sensors, 2023, 23(16): 7112.

[10] S. A. A. Qadri, N. -F. Huang, T. M. Wani and S. A. Bhat, "Plant Disease Detection and Segmentation using End-to-End YOLOv8: A Comprehensive Approach," 2023 IEEE 13th International Conference on Control System, Computing and Engineering (ICCSCE), Penang, Malaysia, 2023, pp. 155-160, doi:10.1109/ICCSCE58721.2023.10237169.

[11] T. Hu, D. Zhuang, J. Qiu and L. Zheng, "Improved YOLOv8 Algorithm with C2f-DCNv3 and Shuffle Attention for Detection of Coal Shearer Drum Teeth," 2024 4th International Conference on Neural Networks, Information and Communication Engineering (NNICE), Guangzhou, China, 2024, pp. 1019-1022, doi: 10.1109/NNICE61279.2024.10498575.

[12] Z. Li, C. Jiang and Z. Li, "An Insulator Location and Defect Detection Method Based on Improved YOLOv8," in IEEE Access, vol. 12, pp. 106781-106792, 2024, doi: 10.1109/ACCESS.2024.3436919.

[13] L. Nannan and W. Liuai, "Research on Helmet Wearing Detection Based on Improved YOLOv8 Algorithm," 2023 IEEE 11th Joint International Information Technology and Artificial Intelligence Conference (ITAIC), Chongqing, China, 2023, pp. 910-917, doi: 10.1109/ITAIC58329.2023.10408800.

[14] X. Hu, P. Zhang, Q. Zhang and F. Yuan, "GLSANet: Global-Local Self-Attention Network for Remote Sensing Image Semantic Segmentation," in IEEE Geoscience and Remote Sensing Letters, vol. 20, pp. 1-5, 2023, Art no. 6000105, doi: 10.1109/LGRS.2023.3235117.

[15] Y. G. Aadarsh and G. Singh, "Comparing UNet, UNet++, FPN, PAN and Deeplabv3+ for Gastrointestinal Tract Disease Detection," 2023 International Conference on Evolutionary Algorithms and Soft Computing Techniques (EASCT), Bengaluru, India, 2023, pp. 1-7, doi: 10.1109/EASCT59475.2023.10392883.

[16] Xiaohui Liu, Yuanfu Zhao, Liyan Liu, Xiaofeng Jin, Chunfang Wang and Yue Zhao, "The analyze and design of low FPN double delta sampling circuit for CMOS image sensor," 2013 IEEE International Conference of Electron Devices and Solid-state Circuits, Hong Kong, 2013, pp. 1-2, doi: 10.1109/EDSSC.2013.6628186.

[17] WU Y, TANG S, ZHANG S, et al. An enhanced feature pyramid object detection network for autonomous driving [J]. 2019, 9(20): 4363.

[18] TAN M, PANG R, LE Q V, et al. Efficientdet: scalable and efficient object detection [C]. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, June 13–19,2020, pp. 10781–10790.

[19] J. Ou and Y. Shen, "Underwater Target Detection Based on Improved YOLOv7 Algorithm With BiFusion Neck Structure and MPDIoU Loss Function," in IEEE Access, vol. 12, pp. 105165-105177, 2024, doi: 10.1109/ACCESS.2024.3436073.

[20] Y. Su, B. Cheng and Y. Cai, "Detection and Recognition of Traditional Chinese Medicine Slice Based on YOLOv8," 2023 IEEE 6th International Conference on Electronic Information and Communication Technology (ICEICT), Qingdao, China, 2023, pp. 214-217, doi: 10.1109/ICEICT57916.2023.10245026.