This repository hosts the AI-based Pesticide Detection and Recommender System, a cutting-edge solution designed to address the critical issue of pesticide overuse in agriculture. Utilizing Python and Flask, this project leverages trained machine learning models to automate pest detection through drone and image analysis, thereby facilitating the targeted application of pesticides. Our system not only identifies pest threats but also recommends the most effective and environmentally friendly pesticides, aiming to significantly reduce pesticide misuse and its associated health and environmental risks.
In Lebanon, the excessive use of pesticides has long been a public health concern and a primary source of water and soil contamination. The prevalent approach involves widespread and unorganized spraying, often involving banned or obsolete pesticides, rather than strategic application. This project seeks to provide a sustainable alternative by automating pest detection and offering tailored pesticide recommendations to promote responsible pesticide use, safeguard public health, and protect local ecosystems.
For a comprehensive understanding of the project's background, development, and testing, please refer to our detailed report: AI-Based Pesticide Detection and Recommender System Report.
To set up the project environment:
git clone https://github.com/Ahmad-ElHajj/eece490_project.git
cd eece490_projectAfter installation, run the Flask application with:
python -m flask --app app --debug runNavigate to the provided local URL to access the web interface, where users can upload images for pest detection and receive pesticide recommendations.
- Automated Pest Detection: Utilizes a Convolutional Neural Network (CNN) model trained on a comprehensive dataset for pest identification from images.
- Pesticide Recommendation: Employs an item-based recommender system considering various metrics (e.g., effectiveness, environmental impact) to suggest the most suitable pesticides.
- User-Friendly Interface: Designed with a simple, intuitive interface for easy submission of pest images and review of recommendations.