Design and Development of a Low-Cost Automated Parking System for Developing Countries

Abstract

Vertical parking systems are becoming popular all over the world, specifically in congested areas because the parking spaces can be maximized by utilizing vertical space with limited land availability. Rapid spread of sensing technology’s also facilitated the development of smart vertical parking technologies. In this work, we propose a low cost rotating smart-parking system by offering a small prototype. The prototype is designed in such a way that it can accommodate six automobiles in a space of two. An Arduino UNO microcontroller with several sensors are used to build the prototype. For vehicle detection, in our case sedan and SUV are considered as standard object, Resnet34 image classification model is used as it provides higheraccuracy according to our experiment results. The objective of this paper is to propose a low-cost automated parking system that can be implemented in existing parking facilities. The prototype performance is measured with object recognition accuracy and system response time.