PLANT DISEASE DETECTION
The SVM, Random Forest classifier, and CNN are found to be efficient and accurate methods of classification and the potential for automation and smartphone-assisted diagnosis is covered.
Abstract
This review paper contains the analysis of the plant disease detection papers on using Convolutional Neural Networks (CNN), Support Vector Machines (SVM), Back Propagation, and K-means Clustering for detecting disease in plants and covers the most recent developments in these techniques. A very famous Plant Village Dataset is chosen and discussed on its data preprocessing and feature selection. The SVM, Random Forest classifier, and CNN are found to be efficient and accurate methods of classification. The review also covers the potential for automation and smartphone-assisted diagnosis as well as suggestions for future research.