International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 05 | May 2025
p-ISSN: 2395-0072
www.irjet.net
SmartAgriDoc: Offline Plant Disease Detection with AI-Powered Mobile App C. Aiman Sulthana¹, Niha Dodamani², Isra Tinmekar³, Prof. Vaibhav Chavan´ ¹B.E. Student, Dept. of Artificial Intelligence and Data Science, Angadi Institute of Technology and Management, Belagavi, Karnataka, India ²B.E. Student, Dept. of Artificial Intelligence and Data Science, Angadi Institute of Technology and Management, Belagavi, Karnataka, India ³B.E. Student, Dept. of Artificial Intelligence and Data Science, Angadi Institute of Technology and Management, Belagavi, Karnataka, India ⁴Professor & Project Guide, Dept. of Artificial Intelligence and Data Science, Angadi Institute of Technology and Management, Belagavi, Karnataka, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Plant diseases cause significant losses in
agriculture, threatening food security and farmer livelihoods. Traditional disease identification methods require expert knowledge, which is often inaccessible to rural farmers. To address this challenge, we propose SmartAgriDoc, an AIpowered mobile application that enables offline plant disease detection using deep learning. The app leverages Convolutional Neural Networks (CNNs) trained on a diverse dataset of healthy and diseased plant leaf images. Integrated with TensorFlow Lite and developed using Flutter, the app offers real-time disease diagnosis and remedy suggestions without the need for internet connectivity. This offline capability ensures accessibility in remote farming regions. Key features include image preprocessing (resizing, normalization), high-accuracy predictions, and a user-friendly interface. Our solution aims to empower farmers with instant and reliable plant health diagnostics, improving early intervention and crop productivity. Experimental results and literature comparisons indicate high accuracy and practicality of the system for real-world deployment. Key Words: Plant Disease Detection, Deep Learning, Convolutional Neural Networks (CNN), Mobile Application, Offline Diagnosis, Smart Agriculture, TensorFlow Lite, Flutter.
agriculture. In particular, Convolutional Neural Networks (CNNs) have demonstrated significant success in image classification tasks, including plant disease detection. Several studies have shown that AI models can surpass human experts in identifying visual symptoms of crop diseases from leaf images. However, most existing solutions rely on cloudbased processing, requiring constant internet connectivity, which limits their applicability in rural environments. To overcome this challenge, we propose SmartAgriDoc, a mobile application that performs plant disease detection in an offline environment using an AI model embedded within the app via TensorFlow Lite. The application is developed using Flutter for cross-platform compatibility and is designed to provide a seamless and intuitive user experience. Farmers can simply capture an image of the diseased leaf using their smartphone, and the app will analyze the image and provide the predicted disease along with suggested remedies—without needing an internet connection.This paper presents the development, architecture, and performance evaluation of SmartAgriDoc. We also compare our work with existing literature and highlight the real-world benefits of our offline AI-powered solution for sustainable agriculture.
1.INTRODUCTION Agriculture plays a vital role in the Indian economy, with a large portion of the population depending on it for livelihood. However, the productivity and quality of crops are often threatened by plant diseases, which can lead to severe economic losses and reduced food security. Early and accurate detection of plant diseases is essential for effective crop management, but traditional diagnostic methods require agricultural expertise and laboratory support, which are often inaccessible to small-scale farmers in remote areas. Recent advancements in artificial intelligence (AI) and deep learning have opened new possibilities in precision
© 2025, IRJET
|
Impact Factor value: 8.315 |
Figure 1- Leaves having visible abnormality
ISO 9001:2008 Certified Journal
|
Page 1565