Research Article
Detecting Citrus in Orchard Environment by Using Improved YOLOv4
Table 4
Kumquat training parameters of four methods in different periods.
| Model | Growth stage | F1 score (%) | Average time/s | Accuracy (%) |
| Faster-RCNN | Growing period | 86.24 | 0.32 | 0.84 | YOLOv3 | Growing period | 83.97 | 0.21 | 0.82 | YOLOv4 | Growing period | 87.51 | 0.12 | 0.92 | Improved YOLOv4 | Growing period | 91.95 | 0.10 | 0.95 | Faster-RCNN | Maturity | 86.48 | 0.32 | 0.86 | YOLOv3 | Maturity | 84.17 | 0.21 | 0.82 | YOLOv4 | Maturity | 87.58 | 0.11 | 0.92 | Improved YOLOv4 | Maturity | 92.13 | 0.09 | 0.96 |
|
|