Table 2 Comparative evaluation of different research works pertaining to paddy leaf disease.
From: An automated hybrid deep learning framework for paddy leaf disease identification and classification
Refs | Problem Addressed | Solution | Benefits | Limitations |
|---|---|---|---|---|
Identification of leaf diseases across multiple crops, including corn, rice, and wheat | DenseNet201 + SVM | 99.82%, 98.75%, 84.15% | Accuracy needs to be improved | |
Detection of cucumber leaf diseases and pests using deep learning | YOLOv51 model | 80.10% | The model architecture requires enhancement to improve the system’s performance | |
Paddy plant leaf disease classification | DL/MBi-LSTM/ SV-RFE and ARFA | 97.16% | This technique is limited to diagnosing certain crop diseases and may not apply universally | |
Automated categorization of multi-class leaf diseases in tomatoes | Deep multi-level convolutional neural network (DMCNN) | 99.10% | Practical deployment in precision agriculture remains untested | |
Identification of Multiple Diseases in Apple Leaf | RegNet DCNN | 93.85% & 99.23% | The accuracy achieved is 93.85%, with 24 samples misclassified | |
Rice leaf disease prediction | IBS-optimized DGAN | 98.70% | Multiclass and a more comprehensive range of leaf features must be considered | |
Automated detection of blast disease in paddy crop | AlexNet, LeNet, VGG 16 | 98.7%, 98.2%, 97.8% | Real-time testing has to be conducted | |
Paddy plant leaf disease classification | SV-RFE and ARO and ABi-LSTM | 98.86% | It suffers from generalizability | |
Detection of brown spot rice leaf disease | CNN and Visual Geometry Group (VGG)19 | 93.00% | The proposed method focuses on a single disease in paddy crops. However, it should also be expanded to cover other crops and improve accuracy | |
Multiclass paddy disease detection | ML (KNN, Random Forest, LDA, Histogram Gradient Boosting) | 90% | Transformer models have yet to be investigated for other crops | |
Detection of Rice Plant Diseases | Deep Convolutional Neural Network (DCNN) | 96.08% | Lack of validation is still a question | |
Rice leaf disease identification | ResNet50 plus SVM | F1-score 98.38% | Focused on a narrow range of rice diseases | |
Early disease detection in rice paddy | IoT-based intelligent farming using CNN | 97.70% | Severity estimation remains a challenge | |
Paddy plant leaf disease classification | GCL | 97% | Severity estimation remains a challenge | |
Fungal disease detection across multiple crops | Modified ResNeXt CNN | 98.92% | Severity estimation remains a challenge | |
Mango crop maturity estimation | Meta-Learning with DenseNet-121 Architecture | 83.65% | Multi-level estimation remains a challenge in harvesting and requires improvements in accuracy | |
Potato leaf disease detection | Data Augmentation using VGG-19 | 99.2% | Severity estimation remains a challenge | |
Diagnosis of Fungal Diseases in Apple Crops | Enhanced ResNeXt Architecture | 98.94% | Severity estimation remains a challenge, and there is a need to adapt the model to other crop types | |
Crop Disease Detection in Agriculture | AI Models Including SVM, ANN, and CNN | 99% | Early-stage severity assessment continues to be a challenge |
