A scalable vision-based smart parking system designed to detect parking slot availability in real-time using computer vision and deep learning. The system eliminates the need for expensive sensor-based solutions by leveraging CCTV/IP camera feeds, YOLO-based detection models, and Flask + Streamlit for backend and dashboard integration.
The architecture includes:
- Flask backend for processing detection results.
- Streamlit dashboard for a real-time user-friendly interface for both administrators and drivers.
- MongoDB for storing parking data, enabling occupancy tracking, usage analysis, and future optimizations.
Planned enhancements include:
- Cloud integration for distributed scalability.
- Predictive analytics for demand forecasting.
- Mobile-based reservations and automated billing.
By reducing search time, minimizing congestion, and optimizing parking utilization, the system contributes to sustainable and intelligent urban mobility, aligning with smart city visions.
- Real-Time Detection: YOLOv8-based model to detect vehicles and parking slot occupancy.
- Web Dashboard: Built with Streamlit for real-time monitoring and analytics.
- Flask API Backend: Processes image/video streams and returns detection results.
- Scalable Storage: Parking data stored in MongoDB for analytics and optimization.
- Cost-Effective: Eliminates the need for physical sensors by using camera-based detection.
- Future-Ready: Supports cloud deployment, predictive analytics, and mobile apps.