Parking Management by Means of Computer Vision

Abstract

Car park management currently relies on parking attendants to show where empty parking lots are by monitoring all available parking areas. In some shopping centers sometimes do not have information about parking conditions, whether parking is full or empty. This condition causes the driver to have difficulty finding a parking space. When drivers are looking for a parking space, many cars are milling about causing congestion and wasting time. This paper presents a breakthrough to make it easier for drivers to find an empty parking space with builds a car parking monitoring and management system based on computer vision. The system input is a camera that is installed in the center of the parking area. The HAAR Cascade Classifier method detects and counts the cars that are parked and then compared with the available parking slots using. If the number of cars currently parked is less than the slots provided, the parking area will inform you that it is still empty along with the number of available slots. Detection of available parking spaces is based on coordinates that have been determined manually on the camera. A total of five slot areas are determined by this camera. The result of this research shows the system accuracy is 90% for the available car parking.