Parking Availability with a GNSS / vision system
Ezio Malis, Manuel Asselot
Nowadays, finding an available parking space in cities is becoming more and more difficult. Drivers waste a lot of time and fuel turning around searching for a place. Our solution aims to make the process easier.
We propose integrating an application that directs the driver to parking spaces available nearby into portable car-navigation systems. The idea fuses GNSS and computer-vision technologies to provide a reliable and very useful service that saves drivers time and money.
Whenever necessary, users can call up an accurate on-screen map of parking spaces in the vicinity. The occupied spaces are marked with a red symbol, while the empty spaces are marked green. When the system is not sure whether or not a parking space is occupied, the space is marked orange. The user can also visualise the path to the nearest empty space and let the car’s navigation system provide the corresponding directions.
A constant-perception system sends information to a centralised server that monitors existing parking spaces in a given city. In order to have the largest urban coverage, we propose three complementary perception systems to determine whether parking spaces are empty or occupied:
The first system is based on information on users’ positions. All users of the proposed service will send anonymous data on their position to the server when they enter or leave a parking space (see figure 1).
The second system is based on a stationary observation system that monitors parking spaces. We can use existing video-surveillance systems or install our own (see figure 2).
The third system is based on a mobile observation system that monitors parking spaces. We propose embedding a vision system into portable car-navigation systems to analyse the images on board (see figure 3).
These three solutions are needed in order to quickly provide extensive coverage to users. At an early stage, users would not have enough information with just one of these systems. Following mass adoption by drivers, we could reduce the number of perception systems in the application.
We are targeting the mass market of frequent car drivers that waste a lot of time finding a parking space in urban areas. There are about 300 million personal vehicles in Europe, and around 10% are equipped with a portable navigation system.
Recent studies conducted in France have shown that the costs related to searching for parking spaces amount to approximately €70 million per year, which includes elements such as noise, car accidents, and air pollution. Factoring in the 70 million total hours drivers spend on this activity, the number increases €600 million per year. Our solution benefits users by saving them time and money. Communities will also benefit from our system in the form of reduced noise, car accidents, and air pollution.