Precision Agriculture: Classifying Banana Leaf Diseases with Hybrid Deep Learning Models

Abstract

The majority of the people in India dependent on farming to earn a living. As a due to climate change, farmers face various challenges. One of them is a reduction in yield, and one of the causes of that is the development of diseases in the plant. The main economic agricultural activity is a banana plantation, particularly in Asian and African nations. Feature extraction using CNN and SVM was used to identify and classify the banana fruit leaf diseases. The dataset was initially improved, precompiled using Matlab code, and then divided into training and testing sections. During the conduct of this research, the ratio employed to divide the data into training and validation was 80:20. After the CNN was implemented successfully, and the SVM models, the maximum average accuracy measured was 94%. According to this study, the suggested model achieves the automatic right diagnosis of banana leaf diseases and gives a workable method for the detection of crop leaf diseases with high recognition accuracy.