On-Street Parking Spot Detection for Smart Cities

摘要

Car parking in crowded cities is a big problem. Drivers have to do a blind search to find a free on-street parking spot. Blind searching not only causes traffic congestion but also fuel and time consumption. Indoor parking garages have sensors or light systems to point out free spots, unfortunately, an indoor approach is not applicable to the on-street parking problem due to its expensive nature. In our proposed solution, parking spots are monitored with roadside cameras. First, street images taken by roadside cameras are collected to form a dataset. Second, a Convolutional Neural Network (CNN) based on this dataset is built. Then the trained CNN analyzes a new street image to check if there is a free spot or not. In this paper, a mobile application is also developed and presented. Our mobile application takes the user’s request and triggers the corresponding roadside cameras, and then notifies the driver about the available parking spots around the region and offers navigation to the spot upon the driver’s confirmation.