A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning

Wu, Nanhui; Shen, Lifan; Chen, Xiang; Luo, Gaofeng; Huang, Yichun; Wang, Juncheng; Tan, Dapeng; Xu, Weixin

doi:10.3390/app16178780

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning

by

Nanhui Wu

^1,2

,

Lifan Shen

¹,

Xiang Chen

^3,4,

Gaofeng Luo

^3,4,

Yichun Huang

^3,4,

Juncheng Wang

^3,4,

Dapeng Tan

¹

and

Weixin Xu

^1,5,*

¹

College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou 310014, China

²

Quzhou Eco-Industrial Innovation Institute ZJUT, Quzhou 324499, China

³

Zhejiang Taizhou Shenhai Expressway Co., Ltd., Taizhou 317000, China

⁴

Zhejiang Expressway Information Engineering and Technology Co., Ltd., Taizhou 310005, China

⁵

Taizhou Research Institute, Zhejiang University of Technology, Taizhou 318000, China

^*

Author to whom correspondence should be addressed.

Appl. Sci. 2026, 16(17), 8780; https://doi.org/10.3390/app16178780

Submission received: 21 July 2026 / Revised: 27 August 2026 / Accepted: 2 September 2026 / Published: 3 September 2026

(This article belongs to the Section Transportation and Future Mobility)

Download Versions Notes

Abstract

Tunnel traffic surveillance is challenged by uneven illumination, severe occlusion, distant small objects, and long-tailed class distributions, while practical edge deployment imposes strict computational constraints. To address these problems, this paper proposes an integrated object detection, structured compression, and edge-deployment framework based on YOLOv8. A C2f-P3A module is developed to enhance spatial and channel feature representation, while Lite-ASPP is introduced to efficiently incorporate multi-scale contextual information. WIoU v3 is adopted to optimize bounding-box regression, and a dependency-preserving channel-level LAMP strategy is employed to adaptively allocate sparsity across eligible layers. Experimental results on the tunnel-surveillance dataset show that the improved dense model achieves a Precision of 90.49%, a Recall of 75.91%, an mAP50 of 83.13%, and an mAP50:95 of 57.60%, outperforming YOLOv8s by 5.46, 6.56, 4.98, and 2.17 percentage points, respectively. Furthermore, a target channel sparsity of 60% reduces the parameter count from 11.24 M to 9.86 M and the computational cost from 28.4 to 25.7 GFLOPs, while retaining an mAP50 of 83.11%. After mixed-precision deployment on the RK3588 platform, the pruned model achieves an mAP50 of 72.87%, an average processing time of 39.04 ms/frame, and a processing rate of 25.61 FPS. These results indicate a favorable empirical trade-off among detection accuracy, model complexity, and edge-processing efficiency under the evaluated conditions.

Keywords: tunnel traffic; object detection; YOLOv8; feature enhancement; adaptive pruning; edge deployment

Share and Cite

MDPI and ACS Style

Wu, N.; Shen, L.; Chen, X.; Luo, G.; Huang, Y.; Wang, J.; Tan, D.; Xu, W. A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning. Appl. Sci. 2026, 16, 8780. https://doi.org/10.3390/app16178780

AMA Style

Wu N, Shen L, Chen X, Luo G, Huang Y, Wang J, Tan D, Xu W. A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning. Applied Sciences. 2026; 16(17):8780. https://doi.org/10.3390/app16178780

Chicago/Turabian Style

Wu, Nanhui, Lifan Shen, Xiang Chen, Gaofeng Luo, Yichun Huang, Juncheng Wang, Dapeng Tan, and Weixin Xu. 2026. "A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning" Applied Sciences 16, no. 17: 8780. https://doi.org/10.3390/app16178780

APA Style

Wu, N., Shen, L., Chen, X., Luo, G., Huang, Y., Wang, J., Tan, D., & Xu, W. (2026). A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning. Applied Sciences, 16(17), 8780. https://doi.org/10.3390/app16178780

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.

Article Menu

A Lightweight YOLOv8 Tunnel Traffic Object Detection Method Based on Feature Enhancement and Adaptive Pruning

Abstract

Share and Cite

Article Metrics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI