Urban areas, university campuses, event centers, and shopping malls frequently struggle with parking congestion. Traditional parking management systems are often inefficient, causing delays, frustration, and increased traffic. We developed Smart Park, an intelligent, camera-based parking management system that leverages computer vision and machine learning to enhance parking efficiency.
We were responsible for:
- Designing and building the system using a microcontroller to handle computing and a camera module to monitor parking lots in real-time
- Implementing YOLOv8 computer vision model to accurately detect available and occupied parking spaces
- Developing a mobile application for users to check parking availability before arriving
- Creating automated and manual parking space mapping for complex and simple lots
- Ensuring power-efficient design capable of running on battery or direct power
- Testing the system rigorously across multiple lots under various environmental conditions
Process & Challenges
- Processing live video footage on the microcontroller while maintaining real-time performance
- Ensuring high detection accuracy under different lighting and weather conditions
- Integrating data transmission to the mobile app for live user updates
- Designing a system that is scalable and reliable for multiple parking lots simultaneously
Outcome & Impact
- Reduced time for users to find available parking spaces, improving traffic flow
- High detection accuracy of parked vehicles across tested lots
- Mobile app provides real-time parking availability, enhancing user experience
- Power-efficient design allows for flexible deployment without complex infrastructure
