The accuracy of road object detection is crucial for ensuring the safe operation of autonomous vehicles. However, existing road object detection models suffer from problems such as missed detection of small objects, excessive parameters, and low accuracy. To address these issues, the present study proposes a road object detection algorithm based on YOLOv8n. First, to reduce model parameters and computational complexity, a lightweight C2f-FB module is constructed by combining C2f and Faster Block to replace the original C2f in the backbone network. Second, lightweight depthwise separable convolution is introduced into the neck network for downsampling, further reducing the number of parameters and computations. Finally, to improve the detection performance of small road objects, a high-resolution branch and processing detection head are added for small object feature extraction. Comparative experiments on the KITTI dataset evaluated the improved algorithm compared with mainstream methods, the object detection accuracy reached 0.901 and 0.634 (in terms of mAP@50 and mAP@50-95), which is a 2.7% and 3.3% improvement compared to the baseline model, while the number of parameters was reduced by 40.95%. The results demonstrate that the model achieves high detection accuracy and a lightweight design, highlighting its reliability and effectiveness in complex traffic scenarios.

Changchun Wang, Yibo Yin, Dongmei Liu. Road Object Detection Algorithm Based on Improved YOLOv8n. Academic Journal of Computing & Information Science (2026), Vol. 9, Issue 4: 49-55. https://doi.org/10.25236/AJCIS.2026.090406.

[1] Wang H, Liu C, Cai Y, et al. YOLOv8-QSD: An improved small object detection algorithm for autonomous vehicles based on YOLOv8[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1-16.

[2] Natan O, Miura J. DeepIPC: Deeply integrated perception and control for an autonomous vehicle in real environments[J]. IEEE Access, 2024, 12: 49590-49601.

[3] Kang L, Lu Z, Meng L, et al. YOLO-FA: Type-1 fuzzy attention based YOLO detector for vehicle detection[J]. Expert Systems with Applications, 2024, 237: 121209.

[4] Chen F, Gu X, Gao L, et al. Pedestrian detection method based on fcos-defpn model[J]. IEEE Access, 2024, 12: 144337-144349.

[5] Zhu J, Liu Y, Zhang D. A Lightweight Road Crack Detection Model Based on Improved YOLOv8n[J]. Journal of Data Acquisition and Processing, 2025, 40(05): 1333-1347.

[6] Huang C, Xu H, Zhang X, et al. BGR-YOLO: An Improved Object Detection AlgorithmUnder Traffic Scenarios Based On YOLOv8[J/OL]. Computer Engineering & Science, 2025: 1-13.

[7] Sang J, Zhang Z, Xiao C, et al. An improved YOLOv8s method and its applicationin road traffic target detection[J]. Infrared and Laser Engineering, 2024, 53(11): 303-316.

[8] Zhu Y, Zhang Y. SEP-YOLO: ROAD OBJECT DETECTION ALGORITHM IMPROVED BASEDON YOLOV8[J/OL]. Computer Applications and Software, 2024: 1-8.

[9] Gao D, Chen T,  Miao L. Improved Road Object Detection Algorithm for YOLOv8n[J]. Computer Engineering and Applications, 2024, 60(16): 186-197.

[10] Liu Z, Li L, Hu Y, et al. Enhanced road object detection with DFPD-YOLO: focusing on small and occluded targets [J]. The Journal of Supercomputing, 2025, 81(14): 1329.

[11]  Chen H, Chen Z, Yu H. Enhanced YOLOv5: an efficient road object detection method[J]. Sensors, 2023, 23(20): 8355.

[12] ULTRALYTICS.YOLOv8. https://github.com/ultralytics/ultralytics.

[13] Chen J, Kao S, He H, et al. Run, don't walk: chasing higher FLOPS for faster neural networks[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2023: 12021-12031.

[14] Chollet F. Xception: Deep learning with depthwise separable convolutions[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 1251-1258.

[15] JIANG Jian, YUAN Zhi qun, GAO Xiu jing, HE Hong zheng, GU Zi-shuo. Object detection algorithm for smart port based on improved YOLOX-S[J]. Object detection algorithm for smart port based on improved YOLOX-S, 2025,46(07):2045-2053.