Plant Disease Detection Using Combined Deep Learning and Machine Learning Techniques

Abstract Effective disease identification in plants is param bount for the success of farming systems. Traditionally, farmers rely on naked-eye observations to recognize disease symptoms in plants, necessitating continuous monitoring. However, this approach becomes cost-prohibitive in large plantations and may be prone to inaccuracies. In certain regions, farmers may need to consult experts by presenting specimens, resulting in time-consuming, less efficient and expensive processes. This study presents a comparative analysis of machine learning and deep learning techniques for automated plant disease detection. Convolutional Neural Network (CNN), Random Forest Algorithm (RFA), and Support Vector Regression (SVR) classification approaches were implemented and evaluated. The proposed system was integrated into a Flask-based web application that enables users to upload plant leaf images for disease classification. Experimental results reveal that the CNN model achieved the highest performance with an accuracy of 82.68%, outperforming Random Forest (74.46%) and SVR (26.84%). Additional evaluation metrics obtained for the CNN model include Precision (0.8610), Recall (0.8442), and F1-Score (0.8417). The findings demonstrate that deep learning techniques, particularly CNNs, provide superior capability for plant disease identification and can contribute significantly to precision agriculture. Keywords: Plant Disease Detection, Deep Learning, Machine Learning, Precision Agriculture, Disease Classification.