Table 2 FW prediction multiple linear regression models with different parameter number (PN) for top only and top and three side view 3D images
From: A method for phenotyping lettuce volume and structure from 3D images
PN | R2 | RMSE, g | NRMSE, % | FW |
|---|---|---|---|---|
Top and three side view 3D images | ||||
1 | 0.85 | 18.2 | 10.1% | 27.2 + 51.8·V |
0.74 | 23.8 | 13.2% | − 149.6 + 30.7·Height | |
0.08 | 44.6 | 24.7% | 74.5 + 43.0·Amp | |
2 | 0.86 | 17.6 | 9.7% | 57.8 + 54.3·V-16.2·Amp |
0.85 | 17.9 | 9.9% | − 6.9 + 44.3·V + 5.2·Height | |
0.74 | 23.7 | 13.1% | − 142.1 + 31.2·Height-5.4·Amp | |
3 | 0.86 | 17.3 | 9.6% | 23.7 + 46.8·V + 5.2·Height-16.2·Amp |
Top view 3D images | ||||
1 | 0.79 | 18.8 | 10.9% | 20.7 + 66.9·V |
0.62 | 25.3 | 14.6% | − 125.2 + 28.1·Height | |
0.04 | 40.4 | 23.2% | 91.1 + 31.5·Amp | |
2 | 0.8 | 18.2 | 10.5% | 58.3 + 70.2·V-18.2·Amp |
0.8 | 18.5 | 10.7% | − 14.5 + 57.4·V + 5.4·Height | |
0.63 | 25.0 | 14.4% | − 157.0 + 27.6·Height + 15.1·Amp | |
3 | 0.81 | 18.1 | 10.4% | 30.5 + 63.7·V + 3.4·Height-15.6·Amp |