Image-based Plant Disease Classification for the Management of Crop Health

Abstract

The classification of plant diseases is essential for ensuring agricultural production and food security. In this research, we look into two distinct methods for classifying plant diseases: Convolutional neural networks for deep learning and logistic regression (LR) for machine learning and RandomForestClassifier(RFC). Using a collection of plant pictures that represent different diseases, we train and assess LR and CNN models. The CNN model automatically learns hierarchical representations, whereas the LR model uses manually created features retrieved from the images. Our analysis indicates that both LR and CNN models can classify plant diseases with high accuracy, with CNN outperforming LR due to its capacity to recognize complicated picture patterns. The conclusions drawn from the results of this experiment show how effective machine learning and deep learning approaches are in grouping plant diseases.