Fig 1.
Schematic workflow of proposed GL-CNN-based model.
Fig 2.
Proposed GL-CNN architecture.
Table 1.
GL-CNN configurations.
Fig 3.
GAP layer vs. FC layer.
Fig 4.
Activation analysis of tanh and ReLU.
Fig 5.
Data augmentation process.
Fig 6.
Sample palm tree plantlings.
Fig 7.
Healthy seedling growth.
Fig 8.
Poor seedling growth.
Fig 9.
IoT monitoring module.
Fig 10.
Preprocessed palm plantlings dataset images.
Table 2.
Accuracy and loss.
Fig 11.
Growth prediction.
Fig 12.
Accuracy with 250 epochs.
Fig 13.
Mean absolute error over 250 epochs.
Fig 14.
Prediction accuracy.
Fig 15.
Prediction performance analysis.
Table 3.
Performance of GL-CNN vs ground truth.
Table 4.
Performance of GL-CNN vs existing models.