Design and Implementation of a Parking Lot Monitoring System Using the YOLOv5 Method (Case Study in the AH Polinema Building Parking Lot)

Abstract

High vehicle productivity and not accompanied by adequate road construction can result in traffic jams on the highway. Apart from causing traffic jams on the roads, the large number of private vehicles, especially four-wheeled vehicles, results in full parking spaces in public service areas. The size of parking spaces in public places cannot keep up with developments in the number of private cars, so that many car parking spaces are full and it is difficult to know which parking positions can still be occupied. The Empty Parking Lot Monitoring System in the AH Polinema Building was carried out based on the level of difficulty in finding empty slots for parking four-wheeled vehicles or cars, so this system was formed as a solution by becoming an information system for the general public to make it easier to find empty slots for parking without visiting the parking lot. This system utilizes two web cameras facing the parking lot of the AH Polinema Building. The captured images from the two web cameras will be processed via a single-board computer (SBC), namely the Raspberry Pi 4B. Image processing was carried out using the YOLOv5 method. This processing results in the calculated value of the total empty slots for the AH Polinema Building parking lot. This output is integrated into the website so that the public can easily access information on vacant parking lots in the AH Polinema Building.