|
| Crop culture | Method used | Performance metrics | Dataset | Accuracy (%) | Reference |
|
| 25 different plants (apple, grape, banana, cherry, etc.) | CNN-AlexNet, AlexNetOWTBn, GoogLeNet, OverFeat, and VGG | Success rate | Open dataset | 99.53 (VGG) | [8] |
| Multicrops (wheat, barley, corn, rice, and rapeseed) | ResNet50 with contextual information | Accuracy | Real-field images | 98.0 | [16] |
| Tomato | Fast R-CNN & Mask R-CNN with VGG16, ResNet50, ResNet101, and MobileNet | Average precision | Internet images | 99.64 (ResNet101) | [5] |
| Multiple plants (like apple, grape, corn, tomato, etc.) | CNN | Accuracy | PlantVillage dataset | 96.5 | [18] |
| Arabidopsis plants | Shallow CNN & Canny edge detector | The difference in count (DIC), foreground/background dice (FBD), symmetric best dice (SBD) | Aberystwyth leaf evaluation dataset (ALED) | 95 | [19] |
| Tomato leaves | PCA-whale optimization & DNN | Accuracy and loss rate | PlantVillage dataset | 86 | [21] |
| Tomato leaves | ResNet50 | Accuracy | PlantVillage dataset | 97 | [22] |
| Tomato leaves | CNN-AlexNet, SqueezeNet, Inception V3, and SVM | Accuracy, recall, F1-score | PlantVillage dataset | AlexNet-93.40, SqueezeNet-90.76, Inception V3-90.43 | [23] |
| Tomato leaves | ResNet and U-Net | Accuracy | PlantVillage dataset | 94 | [24] |
| Tomato leaves | CNN with an attention mechanism | Accuracy | PlantVillage dataset | 98 | [25] |
| Tomato leaves | CNN models—VGG16, VGG19, ResNet, Inception V3 | Accuracy, precision, recall, F1-score | Laboratory & field datasets | (Inception V3) 99.6 & 93.6 | [26] |
| Multiple common disease types | GoogLeNet, VGG16, Inception V3 | Accuracy | PlantVillage dataset | 98 (VGG16) | [28] |
| 14 different plant species | EfficientNet | Accuracy, sensitivity, specificity, precisions | PlantVillage dataset | 98.4 | [29] |
| Tomato leaves | CNN with 4 hidden layers | Accuracy, precision, recall, F1-score | PlantVillage dataset | Average 91.2 | [30] |
| Tomato leaves | CNN with KijaniNet (modified segmentation network) | Mean accuracy, mean boundary F1-score | Real conditioned dataset | 98.46 | [31] |
| Grape leaves | Enhanced ANN and CNN | Accuracy and F1-score | PlantVillage dataset | 93.75 | [32] |
| Tomato leaves | Region-based CNN (R-CNN) | Average precision, confusion matrix | Real-field images | 83.06 | [33] |
| Banana leaves | CNN with a total generalized variation fuzzy C-means segmentation | Sensitivity, specificity, accuracy | Real-field images | 93.45 | [34] |
| Maize plant leaves | CNN-AlexNet | Accuracy | PlantVillage dataset | 99.16 | [35] |
| Mix crop leaves (wheat, corn, cotton, grape, and cucumber) | CNN-AlexNet with PSO optimization | Sensitivity, specificity, accuracy, precision, F1-score | Real-field images | 98.83 | [36] |
|
