Smart Crop Disease Using CNN Model

Agriculture continues to face significant challenges due to crop diseases that result in reduced yield, economic losses, and delayed intervention, particularly in developing regions where access to expert diagnosis is limited. Traditional disease identification methods rely on manual inspection, which is time-consuming, subjective, and not scalable. This paper presents a Smart Crop Disease Detection System using Convolutional Neural Networks (CNNs) for automated and accurate identification of plant diseases from leaf images. The proposed system leverages deep learning techniques trained on real-world agricultural image data obtained from the PlantDoc dataset, which contains healthy and diseased crop leaves captured under diverse field conditions. A lightweight and efficient CNN architecture, MobileNetV2, is adopted to enable real-time disease detection with reduced computational overhead, making the system suitable for mobile and low-power devices. The model performs image classification to identify disease categories and assess plant health conditions. Experimental evaluation demonstrates that the proposed model achieves an accuracy of 85%, outperforming other baseline architectures. To enhance deployability, the trained model is converted into TensorFlow Lite, enabling seamless integration into mobile and web-based applications. The proposed framework facilitates early disease detection, supports timely preventive measures, and contributes to improved agricultural productivity through intelligent decision support.