An Improved YOLOv11-Based Model for Pavement Defect Detection.

Saved in:

Bibliographic Details
Title:	An Improved YOLOv11-Based Model for Pavement Defect Detection.
Authors:	Han, Jiatong¹ 18741931949@163.com, Wang, Li² wangli9966@ustl.edu.cn
Source:	Engineering Letters. Jul2026, Vol. 34 Issue 7, p2570-2583. 14p.
Subjects:	Feature extraction, Object recognition (Computer vision)
Abstract:	To address the issues of large-scale variations and the easy loss of subtle features in complex pavement defect detection, this paper proposes an improved framework based on the YOLOv11 baseline. The SPDConv (Spatial Pyramid Depthwise Convolution) module is introduced into the backbone network to enhance the response to detailed features of multi-morphological defects using multi-scale convolutions. A lightweight Slim-neck design is adopted in the neck, which reduces computational burden while maintaining feature representation ability. Furthermore, the C2PSA SEAM module is embedded to effectively suppress interference from light reflections and stains. On the RDD2022 China and UAVPDD Computer Vision datasets, the proposed method achieves improvements of 2.6% and 2.3% in mAP@0.5, 2.6% and 6.8% in mAP@0.5:0.95, and 6.7% and 0.2% in recall rate, respectively, verifying the effectiveness and robustness of the approach. [ABSTRACT FROM AUTHOR]
	Copyright of Engineering Letters is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	To address the issues of large-scale variations and the easy loss of subtle features in complex pavement defect detection, this paper proposes an improved framework based on the YOLOv11 baseline. The SPDConv (Spatial Pyramid Depthwise Convolution) module is introduced into the backbone network to enhance the response to detailed features of multi-morphological defects using multi-scale convolutions. A lightweight Slim-neck design is adopted in the neck, which reduces computational burden while maintaining feature representation ability. Furthermore, the C2PSA SEAM module is embedded to effectively suppress interference from light reflections and stains. On the RDD2022 China and UAVPDD Computer Vision datasets, the proposed method achieves improvements of 2.6% and 2.3% in mAP@0.5, 2.6% and 6.8% in mAP@0.5:0.95, and 6.7% and 0.2% in recall rate, respectively, verifying the effectiveness and robustness of the approach. [ABSTRACT FROM AUTHOR]
ISSN:	1816093X