Design Of Deep Learning Model Applied For Smart Parking System
Van An Vo, Van Duc Phan, Vu Minh Bui, Tri Nhut Do
DOI: 10.15598/aeee.v21i4.5366
Abstract
This article proposes and introduces a smart parking system using RFID technology incorporating a Deep Learning model to identify license plates. It tries to simulate the ability of the brain to recognize, differentiate and learn patterns from data. The employed algorithms are mainly based on neural network models where neurons are organized in stacked layers. The system is designed to manage incoming and outgoing vehicles by collecting and processing images and data on passenger information to update parking status with the news of empty lots. Another function of the parking system also designed is a fully automatic method of paying the parking fee by the user. The deep learning model for the smart parking system is implemented using the Raspberry PI 3 embedded system and sensors. Experimental results with the plate identification rate in the worst condition, up to 80%, have proven the reliability of the proposed smart parking system. In terms of quantity, the percentage of the worst plate identification down to 10% has established the stability of the proposed smart parking system.