Table 1 Summary of related works13,14,15,16,20,21,22,23,24,25,26,27,28,29.
Author | Title | Method | Dataset | Accuracy/Result | Limitation |
|---|---|---|---|---|---|
Abeje et al.13 | Sesame disease detection using stepwise deep learning | CNN | 540 images (Ethiopia) | 98% testing accuracy | Small dataset |
Lu et al.20 | Rice disease recognition using Atrous CNN & TL | CNN-SVM | 619 rice leaf images | 91.37% | Limited data; overfitting risk |
Haimanot21 | Sesame disease identification | CNN, AlexNet, GoogLeNet | EIAR (Ethiopia) | GoogLeNet: 95.5% | No preprocessing details |
Eyerusalem22 | Sesame disease classification | Deep CNN + SegNet | Gondar & Humera | 96.67% | Single technique used |
Bashier IH et al.23 | Sesame seed disease detection | CNN (VGG/ResNet) | 1,695 images (Sudan) | Max: 88.5% | Needs improvement |
Tadele AB et al.24 | Step-by-step CNN for sesame diseases | Deep CNN | 540 images (Dejen, Ethiopia) | 98% testing accuracy | Low sampling |
Jasrotia et al14. | Maize disease detection | Optimized CNN | Kaggle | 98.9% | Class imbalance |
Marcos et al.15 | Coffee leaf rust detection | CNN | 159 images | 95% | Small dataset, limited disease range |
El-Mashharawi et al.16 | Tomato disease expert system | CLIPS Rule-Based | - | Field-tested | Low adaptability |
Arivazhagan et al.25 | Mango disease detection | ResNet50 | 8,853 images | 91.5% | No parameter tuning |
Ashqar et al.26 | Tomato disease classification | CNN | - | 99.84% | Not specified |
Gurusamy et al.27 | Potato blackleg detection | ResNet18, ResNet50 | 532 images | Promising | Long training time, no validation |
Apple disease detection | CNN | - | 98.54% | Not detailed | |
Jadhav et al.29 | Soybean disease detection | AlexNet, GoogLeNet | 1,199 images | AlexNet: 98.75% | Limited data |
