轻量级改进RT-DETR的葡萄叶片病害检测算法
|
|
刘慧,王防修,王意,黄淄博,苏晨
|
Lightweight improved RT-DETR algorithm for grape leaf disease detection
|
|
Hui LIU,Fangxiu WANG,Yi WANG,Zibo HUANG,Chen SU
|
|
| 表 6 不同模型的对比实验结果 |
| Tab.6 Comparison experimental result of different models |
|
| 模型 | P/% | R/% | mAP@0.5/% | ${N_{\text{P}}}$/106 | $v$/(帧⋅s−1) | FLOPs/109 | | SSD(2016) | 80.7 | 76.7 | 83.1 | 26.2 | 32.5 | 62.6 | | Faster R-CNN(2016) | 81.6 | 75.3 | 82.8 | 41.3 | 12.9 | 212.8 | | EfficientDet(2020) | 83.7 | 78.6 | 84.5 | 33.4 | 13.4 | 260.7 | | YOLOv7[29](2022) | 89.8 | 88.7 | 92.6 | 36.4 | 82.8 | 103.2 | | YOLOv8s(2023) | 91.0 | 89.5 | 93.0 | 11.1 | 122.3 | 28.4 | | YOLOv9s[30](2024) | 89.4 | 88.1 | 92.1 | 9.6 | 98.7 | 38.7 | | YOLOv10s[31](2024) | 90.6 | 87.7 | 92.5 | 8.1 | 130.2 | 24.5 | | YOLOv11s[32](2024) | 91.2 | 88.6 | 93.1 | 9.4 | 124.2 | 21.3 | | Deformable DETR(2020) | 89.1 | 88.0 | 91.0 | 39.9 | 10.8 | 179.6 | | DINO(2022) | 89.4 | 87.9 | 92.3 | 46.7 | 7.3 | 279.2 | | MS-DETR(2023) | 89.6 | 88.5 | 92.3 | 53.5 | 30.59 | 117.1 | | RT-DETR(2023) | 89.0 | 87.4 | 91.1 | 19.9 | 67.2 | 57.0 | | SCGI-DETR(本文方法) | 91.6 | 89.8 | 93.4 | 10.7 | 66.7 | 20.5 |
|
|
|