湖北汽车工业学院学报2025，Vol.39Issue(1)：15-19,5.DOI:10.3969/j.issn.1008-5483.2025.01.003
基于改进YOLOv8的路面缺陷检测模型
Pavement Defect Detection Model Based on Improved YOLOv8
摘要
Abstract
An improved YOLOv8-based pavement defect detection model was proposed.The C2f_Si-mAM module was designed,and the backbone feature extraction network of YOLOv8 was replaced with GhostNet.Moreover,the SimSPPF module was introduced,and the MPDIOU loss was regarded as the bounding box regression loss.The results indicate that compared to the YOLOv8n model,the improved model achieves a 3.6%increase in average precision,a 10%reduction in the number of parameters,and an FPS of 85,meeting the requirements for real-time detection.关键词
路面缺陷检测/YOLOv8/C2f_SimAM/GhostNetKey words
pavement defect detection/YOLOv8/C2f_SimAM/GhostNet分类
信息技术与安全科学引用本文复制引用
黄配乐,王生怀,陈晓辉,王宸,张伟..基于改进YOLOv8的路面缺陷检测模型[J].湖北汽车工业学院学报,2025,39(1):15-19,5.基金项目
国家自然科学基金(51675167) （51675167）
湖北省重点研发计划项目(2021BAA056) （2021BAA056）
湖北省高等学校优秀中青年科技创新团队计划项目(T2020018) （T2020018）
湖北省自然科学基金(2020CFB755) （2020CFB755）
湖北省教育厅科研项目(Q20191801) （Q20191801）
