基于YOLOv11n的改进铁路工人安全穿戴检测模型

武晓春,李梓宁

Improved railway worker safety wear detection model based on YOLOv11n

Xiaochun WU,Zining LI

表 7 不同模型的检测性能对比结果

Tab.7 Comparison result of detection property of different models

模型 P/% R/% mAP50/% mAP50:95/% FLOPs/109 Np/106 v/(帧·s−1) Faster R-CNN 69.7 54.4 62.5 45.0 948.1 28.3 47 SSD 85.5 84.3 90.8 59.8 6.2 3.8 131 Mamba-YOLO 95.5 90.8 95.7 66.9 13.6 5.98 62 GhostNetP2 93.4 84.9 91.8 56.9 7.3 1.45 167 YOLO-Worldv2 95.3 91.7 95.6 66.9 9.6 3.53 152 RT-DETR 95.7 92.6 96.5 70.5 96.6 25.66 125 YOLOv5n 95.8 90.2 95.4 66.8 7.8 2.18 230 YOLOv8n 94.6 85.5 91.2 57.9 8.1 3.00 278 YOLO11n 94.4 90.3 94.8 65.4 5.5 2.58 197 VMS-YOLO 98.0 92.5 97.5 67.8 5.4 2.45 238