Vision-Based Smart Parking Detection System Using Object Tracking

Parking is essential to all vehicle owners. The inconvenience caused to disabled people due to the lack of disabled parking bays is a real problem, especially in urban areas. This paper proposes a solution consisting of three parts: first, it provides disabled drivers with real-time disabled parking information so they can plan their trips ahead; second, it provides parking operators with an insight into the trends of disabled parking demand so they may provide sufficient disabled parking bays at all times; finally, it prevents non-disabled drivers from parking in the disabled parking bays. The real-time disabled parking information is provided by a Raspberry Pi-based visual tracking system. It tracks vehicles in the disabled parking bays and sends the information to a server. Then, an android application delivers this information to the disabled drivers in real-time. Additionally, a dashboard keeps track of the disabled parking information and presents the usage data history in neat and easy to read graphs so that parking operator can conveniently review the usage trends. To prevent usage abuse, an IoT-enabled barricade is deployed at the center of each disabled parking bay. The barricades can only be controlled by registered disabled drivers. The proposed system provides a resolution to the problem and is an improvement upon existing solutions.