IoT BASED SMART PARKING SYSTEM

Abstract

The increased population has resulted in an unstructured and poorly maintained parking system in the Kathmandu Valley. Smart parking system presents the solution by offering systematic control of the parking area along with suitable billing procedures. The main aim of this project is to design a well-managed parking system with proper billing procedure and security to the user. To achieve the proper implementation of this system, firstly the characters in the number plate of the user’s vehicle is recognized which had been trained using Deep Learning prior the implementation phase. After the recognition process, the recognized characters are then registered into the database with entry time and exit time of that specific vehicle. This data stored in the database helps in calculating the billing amount after the user is done with the parking services. The YOLOv4 object detection model was used for deep learning. The trained model is capable of detecting character like बा, प, च and numeric character from ० to ९. A database is created using SQLite. The detected number plate is stored in database with check-in time and check-out time. The status of parking area can be checked using Adafruit IO. Finally, the project was implemented in Jetson Nano.