🌱 Inspiration Agricultural productivity is often threatened by plant diseases that go undiagnosed until it’s too late. We were inspired by the struggles of small-scale farmers who lack access to expert diagnosis. Our goal was to create an accessible, real-time disease detection tool using just a smartphone camera and artificial intelligence.
🌿 What it does SmartAgro Diagnoser is an AI-powered mobile/web-based system that analyzes leaf images to detect and classify plant diseases in real-time. It tells the user whether a plant is healthy or infected, and if infected, specifies the disease and suggests general remedies or advice.
🛠 How we built it We used a large labeled dataset (like PlantVillage) containing thousands of images of diseased and healthy leaves.
A Convolutional Neural Network (CNN) model was trained using TensorFlow/Keras.
The frontend (web/mobile) allows users to upload or capture a leaf image.
The backend processes the image and feeds it to the trained model to return results instantly.
We deployed the model using Flask and integrated it into a responsive web interface using HTML, CSS, and JavaScript.
🧱 Challenges we ran into Limited access to high-quality, diverse datasets for certain crops.
Balancing model accuracy with size/speed for real-time use on mobile devices.
Making the user interface intuitive for farmers unfamiliar with technology.
Preventing misclassification due to image noise or unclear photos.
🏆 Accomplishments that we're proud of Developed a lightweight yet accurate CNN model with over 90% accuracy.
Created an easy-to-use interface accessible to non-technical users.
Enabled offline image analysis for areas with limited internet access.
Built a tool that could make a real difference in the lives of farmers.
📚 What we learned Deep learning is powerful, but the quality of the dataset is just as important.
Real-world AI applications must consider user accessibility and device limitations.
We learned to integrate AI models into full-stack applications and improve UX for rural users.
🚀 What’s next for SmartAgro Diagnoser: AI System for Plant Disease Detection Expand the dataset to cover more crops and rare diseases.
Add voice-based support in local languages for accessibility.
Partner with agricultural NGOs or governments for large-scale deployment.
Implement geo-tagging and disease tracking features for early warning systems.
Develop a native mobile app for offline usage in low-connectivity areas.
Log in or sign up for Devpost to join the conversation.