Table 3.
Comparison of the outcomes of different tea leaf disease detection algorithms.
| Model name | Diseases | Results | References |
|---|---|---|---|
| Deep CNN (LeNet) |
1. Leaf blight 2. Blight disease 3. Red leaf spot 4. Red scab |
Detection accuracy: 90.23% MCA: 90.16% |
49 |
| CNN |
1. Algae leaf spot 2. Gray Blight 3. White Spot 4. Brown Blight 5. Red Scab 6. Bud Blight 7. Leaf blight |
Detection accuracy: 94.45% | 50 |
| DNN |
1. Gray Blight 2. Algal spot 3. Brown Blight 4. Helopeltis 5. Red Spot |
Detection accuracy: 96.56% Precision: 96.63% Recall: 96.49% F1-score: 0.965 |
51 |
| AX-RetinaNet |
1. Algae leaf spot 2. Bud blight 3. White scab 4. Leaf blight |
mAP: 93.83% Precision: 96.75% Recall: 94.00% F1-score: 0.954 |
7 |
| Improved DCNN |
1. Bud blight 2. Leaf blight 3. Red scab |
Detection accuracy: 92.50% | 4 |
| CNN (LeafNet) |
1. Bird's eye spot 2. Gray Blight 3. White Spot 4. Brown Blight 5. Red leaf spot 6. Algal leaf spot 7. Anthracnose |
Detection accuracy: 90.23% MCA: 90.16% |
52 |
| Multi-objective image segmentation |
1. Red Rust 2. Red Spider 3. Thrips 4. Helopeltis 5. Sunlight Scorching |
Detection accuracy: 83% Precision: 77% Recall: 84% F1-score: 0.780 |
53 |
| Improved YOLOv5 |
1. Tea cell eater (Apolygus lucorum) 2. Leaf blight |
Detection accuracy: 91% Precision: 87.80% Recall: 85.27% mAP: 85.35% FPS: 51 |
55 |
| YOLOv7 |
1. Red spider 2. Tea mosquito bug 3. Black rot 4. Brown blight 5. Leaf rust |
Detection accuracy: 97.30% Precision: 96.70% Recall: 96.40% mAP: 98.2% F1-score: 0.965 |
Present study |