Segmentation And Classification of Plant Leaf Disease Using Advanced Deep Learning Approach and Ensemble Classifier

DOI: https://doie.org/10.0714/Jbse.2024192880

Suma S H , K. Prabhushetty , Jagadish S. Jakati

Keywords:

Plant disease, CNN Mask-RCNN, SVM, ensemble, classification, Machine Learning, Deep Learning.

Abstract:

Plants play a crucial role in sustaining the world's food production. However, the occurrence of various plant diseases poses a threat to crop yields, leading to significant losses if not managed effectively. The traditional approach of manually monitoring plant diseases by agricultural experts and botanists is laborious, demanding, and prone to errors. To mitigate the impact of plant diseases, there is a growing recognition of the potential of machine vision technology combined with artificial intelligence. By automating disease detection and analysis, AI can provide faster and more precise assessments compared to traditional methods. This technological advancement offers a promising solution to reducing the severity of diseases and minimizing crop losses. Recently, deep learning-based methods have gained huge attention from research community in this context of image processing tasks. Therefore, in this work, we present a deep learning enabled ensemble machine learning approach for plant disease classification. The first phase of the work performs data augmentation, in next stage, we present modified Mask RCNN model for plant leaf segmentation. Later, a CNN based model is presented to extract the deep features. Finally, SVM, Random Forest and Decision Tree are used to construct the ensemble classifier with the help of majority voting. The performance of proposed approach is validated on Plant Village, Apple, Maize, and Rice where overall accuracy is obtained as 99.45%, 96.30%, 96.85%, and 98.25%, respectively

PDF