Vacant Car Parks Detection Using Digital Image Processing Methods

Abstract

Long car queues are often encountered in some public facilities because visitors should be around to find an empty parking space. One way to minimize this case is to use a parking information system that shows the location of the parking lot that is empty or occupied with their amounts. This research presented two digital image processing methods for detecting empty space occupied in the image of the car parking area. There are vehicle detection and edge detection method. Vehicle detection is the method used to detect objects in the image by subtracting the parking area image, an empty parking lot, from the image containing the car. In contrast, the edge detection method detects the object's edge. The results from these two methods were then compared using the AND function to obtain the condition of an empty or occupied box for each box in the parking lot. Threshold values affect the determination of the parking lot. In this research, the data used are images of open car parks in the Malang Town Square (Matos) shopping center, Mall Olympic Garden (MOG), and data sourced from journals with similar topics [16]. The test results show that the best detection results are obtained in detecting occupied parking spaces in the parking lot in Malang Town Square (Matos), with a threshold of 10 and an accuracy of 99.4% with a threshold of 10.