Smart Parking System based on Improved OCR Model

Abstract

Smart parking system becomes essential nowadays specially in urban area because it can reduce the time and the fuel wasted in searching for an empty parking slot. The aim of this paper is to develop smart parking system based on improved Optical Character Recognition (OCR) model. The proposed system consists of three stages: 1) OCR based parking slot detection, 2) User notification based on IoT approach, 3) Smart parking meter based on Simple Mail Transfer Protocol (SMTP). In this system, the vacant parking slots are detected using improved OCR model by labeling the parking slots with specific characters. The system identifies empty parking slots by detecting these characters with installed camera above the parking slots, otherwise the parking slots are occupied. The performance of the proposed OCR model in detecting these characters is enhanced by considering two stages of advanced morphology filter to remove unwanted small/large objects from the image. The accuracy of the proposed OCR model is tested by eight images having different parking street textures as a background and characters written with seven font styles. The idea of detecting empty parking spaces is to share the number of vacancies on a website for the drivers searching for a parking space. However, a time-controlled parking system with different charge rate each hour can be considered to make the best utilization of parking spaces available in urban areas. Therefore, a smart parking meter is proposed to notify the user through Email about the time left that allowed to be parked and the change in charge rate. The proposed system is developed and implemented using vision assistant of LabView.