Table 4.

Comparison of detection performance of different models.

Model	Precision (%)	Recall (%)	mAP50 (%)
Faster R-CNN	89.5	86.3	89.3
YOLOv3	85.1	78.9	86.7
YOLOv4	88.7	89.1	89.6
YOLOv5s	93.4	92.1	90.9
YOLOv7-tiny	94.2	88.3	90.1
YOLOv8-n	91.6	85.4	91.8
YOLOv8-ECS	97.2	97.4	98.9