Development of vehicles number plate recognition using optical character recognition technique

Md Nazari, Muhammad Faris (2022) Development of vehicles number plate recognition using optical character recognition technique. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

	Text (24 Pages) Development of vehicles number plate recognition using optical character recognition technique.pdf - Submitted Version Download (600kB)
	Text (Full Text) Development of vehicles number plate recognition using optical character recognition technique.pdf - Submitted Version Restricted to Registered users only Download (6MB)

Abstract

This abstract describes a project aimed at developing a system for recognizing vehicle license plates using optical character recognition (OCR) techniques. The system is intended to automatically detect and recognize the characters on a license plate, and it could have a variety of applications such as traffic management, parking management, security monitoring, and crime prevention. OCR is a technology that can recognize and extract text from an image, it is a process that consist of several stages including pre-processing, segmentation, recognition, and post-processing. The challenges of this project include achieving high accuracy in recognizing license plate numbers, handling variations in license plate styles and environmental conditions, and dealing with moving vehicles. The proposed solution is to use OCR in conjunction with other computer vision techniques such as image processing, pattern recognition, and machine learning. The system is expected to improve the process of recognizing license plates, making it faster and more accurate. Using OpenCV and optical character recognition, this system detects and reads number plates automatically. When a key(s) is pushed on the keyboard, the Pi camera module saves the most recent frame as a new picture while continuing to take frames. Then, it finds the number plate using OpenCV's contour function. Finally, the number plate numbers are read using optical character recognition, which Raspberry Pi does by cropping off that specific region. After collecting the data for the development of a vehicle number plate recognition system using OCR technique, there are some of the key considerations for analysing the data. To evaluate the performance of the developed system, two key considerations are analysed. The first parameter is the accuracy of character recognition which reflects the quality of the images in terms of image clarity and legibility. Out of the 30 sample images tested, 22 samples were properly identified, yielding an accuracy rate of 73.33%. The second parameter is the image classification which reflects the ability of the system to handle different scenarios in terms of the type of vehicles, license plate formats, and backgrounds. Out of the 30 sample images tested, 10 samples were properly identified, yielding a classification accuracy rate of 33.33%. As a result, the goals of this vehicle number plate identification system were fulfilled satisfactorily.

Actions (login required)

View Item