An Intelligent Safety Surveillance System
Smart Eye is an open-source, AI-powered surveillance system that uses computer vision to automate safety monitoring in real time. It combines object detection, face detection, face recognition, and Personal Protective Equipment (PPE) detection into a single, modular application designed for industrial, construction, and other safety-critical environments.
- Object Detection — Detects and localizes people and relevant objects in a video stream.
- Face Detection & Recognition — Identifies and verifies individuals against a known database.
- PPE Detection — Checks whether required safety gear (e.g., helmets, vests) is being worn and flags violations.
- Real-Time Inference — Runs deep learning models through ONNX Runtime for fast, cross-platform performance.
- Modular Architecture — Cleanly separated backend, frontend, and data layers for easy extension and maintenance.
- Automated Testing & CI — GitHub Actions workflows help maintain code quality.
Smart-Eye/
├── .github/workflows/ # CI/CD pipelines
├── backend/ # Detection & recognition pipelines, core logic
├── frontend/ # User interface and visualization
├── data/ # Datasets, models, and supporting data
├── scripts/ # Utility and automation scripts
├── tests/ # Automated tests
├── utils/ # Shared helper modules
├── main.py # Application entry point
├── build.sh # Build script
├── requirements.txt # Core dependencies
└── README.md
- Python 3.9+ (recommended)
pipandvirtualenv
# Clone the repository
git clone https://github.com/ABO47/Smart-Eye.git
cd Smart-Eye
# (Optional) create a virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements.txtpython main.pyFor development, install the additional tooling:
pip install -r requirements-dev.txtpytest tests/| Component | Technology |
|---|---|
| Language | Python |
| Inference Engine | ONNX / ONNX Runtime |
| Computer Vision | Object detection, face recognition, PPE detection |
| CI/CD | GitHub Actions |
| Linting | Ruff |
Contributions are welcome. Please fork the repository, create a feature branch, and open a pull request. Make sure tests pass and code is formatted before submitting.
This project is licensed under the MIT License — see the LICENSE file for details. Third-party licenses are listed in THIRD_PARTY_LICENSES.txt.
Developed and maintained by ABO47.