A Systematic Analysis of Machine Learning and Deep Learning Based Approaches for Plant Leaf Disease Classification: A Review

Table 1

Machine learning-based methods: a comparative analysis.

Crop culture	Methods used	Performance metrics	Dataset	Accuracy (%)	Reference
Tomato plant	Moth-flame optimization & genetic algorithm	Recall, precision, accuracy, F1-score	UCI machine learning repository	86	[10]
Apple & cucumber plant	-means clustering and PHOG	Recognition rate	Real-field images	90.43 & 92.15	[6]
Plant leaves (like rose, lemon, mango, and banana)	-means clustering, genetic algorithm, SVM	Accuracy	Real conditioned capture images	95.71	[11]
Rice plant	Radial basis function neural network	Accuracy, precision, recall	Real-field images	95.0	[12]
Cucumber plant	-means clustering, SVM	Accuracy	Real-field images	86	[7]
Rice crop	KNN, ANN	Accuracy	Real-field images	86 & 99	[2]
Tomato plant	Extreme Learning Machine (ELM)	Accuracy, AUC	Tomato powdery mildew dataset (TPMD)	89.19	[13]
Potato leaves	Capsule networks (CapsNet)	Accuracy	Plant village dataset	91.83	[14]
Tomato plant	SVM & logistic regression (SVM-LR)	Accuracy, AUC, F1-score	Real-time data of tomato powdery mildew disease dataset	92.73	[15]
Rice plant	SVM	Accuracy, sensitivity, specificity, AUC, ROC, F1-score	Real-field images	94.65	[3]