Implementation of an Image Processing based Smart Parking System using Haar-Cascade Method

Abstract

In highly populated cities, finding available car parking slots is time consuming and may cause severe traffic congestions at the parking entrance. Therefore, a smart parking system with automated car detection is required so that the car drivers would have minimum effort and time to access the available parking location. This paper presents an implementation of image processing based smart parking system using Haar-Cascade method. The aim of this project is to develop a smart parking system on a single-board computer and a camera installed at a parking area. The image processing for car detection is performed on Raspberry Pi interfaced with the Firebase cloud platform through an API. The system posts the data to the Firebase channel, enabling an Internet of Things (IoT) based parking system. A mobile application is developed for visualization and hence the data is accessible through the car driver’s phone. The system was tested at a simple parking lot with a line of car slots. The experimental results with different condition of parking areas and different camera view angles are discussed in the paper.