People usually circling around for quite a long time in a large parking garage to find an open parking space due to unevenly distributed available parking spaces. In this paper, we propose a system which uses the convolutional neural network deep learning approach to analysis videos taken at different places in a large parking garage. Our system is able to monitor the availability of each parking space and provide useful information to users about the opening parking spaces in real-time. To evaluate our system, we conduct an empirical analysis based on the data we collected from a parking lot at IKEA in Wuhan. The evaluation results show that our parking space management and allocation system can reduce ineffective vehicle detours, shorten the average time spent on finding an open space. This work also offers useful and deep insights for solving parking problems at centralized and large-scale parking garages.

CICTP