Plant Disease Detection Using Deep Learning And Image Processing

Abstract

Plant diseases pose a significant threat to agricultural productivity, leading to substantial crop losses and economic challenges for farmers. Early and accurate identification of plant diseases is essential to ensure timely treatment and improve crop yield. This project presents a Plant Disease Detection System using Deep Learning and Image Processing that enables automated disease diagnosis from plant leaf images. The proposed system utilizes lightweight deep learning models such as MobileNetV2 and EfficientNet-Lite, optimized through TensorFlow Lite (TFLite) for deployment on Android devices. Leaf images captured through a smartphone camera or selected from the gallery undergo preprocessing techniques including resizing, normalization, and noise reduction before classification. The trained model identifies diseases and provides prediction results along with confidence scores and remedy suggestions. The application is designed to operate offline, making it suitable for farmers in remote areas with limited internet connectivity. Experimental evaluation demonstrates high accuracy, precision, and sensitivity, confirming the effectiveness of the system. The proposed solution supports smart agriculture by providing a fast, cost-effective, and user-friendly tool for real-time plant disease diagnosis and crop management.