Table 1. Summary of related work on plant leaf disease recognition.
| Reference | Methods | Results |
|---|---|---|
| Abdullakasim et al. (2011) | Fully connected neural network (NN) with one hidden layer | 79.23% of diseased leaves, 89.92% of healthy plants (accuracy) |
| Ramcharan et al. (2017) | Inception v3 convolutional neural network (CNN) | 93% (accuracy) |
| Ferentinos (2018) | VGG CNN | 99.53% (accuracy) |
| Ramcharan et al. (2019) | Single Shot Multibox (SSD) model with the MobileNet detector and classifier | 94% ± 5.7% (accuracy) |
| Coulibaly et al. (2019) | VGG16 CNN | 95.00% (accuracy), 90.50% (precision), 94.50% (recall), 91.75% (f1-score) |
| Sangbamrung, Praneetpholkrang & Kanjanawattana (2020) | Custom 15-layer CNN | 0.96 (f-score) |
| Sambasivam & Opiyo (2020) | Contrast Limited Adaptive Histogram Equalization (CLAHE), Synthetic Minority Over-sampling (SMOTE), image flipping and custom 7-layer CNN | 93% (accuracy) |
| Abayomi-Alli et al. (2021) | Color space augmentation and MobileNetV2 CNN | 99.7% (accuracy) |