-
-
NeuraPark
Today's drivers face immense challenges finding parking in urban areas, leading to increased frustration, congestion, and emissions. Traditional parking solutions fall short in providing real-time updates and accurate predictions.
NeuraPark offers a cutting-edge parking management system. By integrating with existing infrastructure such as cameras and parking APIs, it detects real-time parking space availability and predicts lot occupancy for the near future. This ensures drivers can locate parking spots efficiently. The system also allows clients to incorporate sensors and keycard/RFID readers to provide users with information on parking availability and lot traffic. Moreover, it utilizes camera technology for advanced tracking based on number plates, offering insights on unoccupied spots, predicting spot availability, and introducing a flexible payment system.
We developed a prototype using the YOLO object recognition algorithm and the OpenCV computer vision tools library. This system identifies when a car crosses a designated virtual line, marking its entry and exit. Additional cameras focused on car number plates ensure individual vehicle tracking, prediction, and payment integration.
Our algorithm encountered glitches due to limited camera data and processing capabilities, preventing the system from becoming fully functional within our timeline.
Despite being a two-member team, we achieved significant progress in conceptualizing and developing our system. We're proud of our innovative approach to address a modern-day challenge.
This journey acquainted us with powerful tools like OpenCV, YOLO, and the Firebase Cloud database management system. It was a rich learning experience in artificial intelligence and its potential real-world impacts.
We aim to refine our algorithm, address current glitches, and expand on our data sources and processing power to elevate NeuraPark into a comprehensive and user-friendly parking solution.
Log in or sign up for Devpost to join the conversation.