Vehicle Number Plate Identification and Recognition of Number Plate Using Deep Learning Based on Approaches in Images and Videos

Vehicle Number Plate Identification and Recognition of Number Plate Using Deep Learning Based on Approaches in Images and Videos

Manisha Dhargawe1, Prof. R.Y Bhandekar 2, Prof. Rashmi Kannake3, Prof. Monika Ingole4

1 Student, M-Tech Computer Science Engineering Department, WCEM, Nagpur

dhargawemanisha61@gmail.com

2,3,4Assistant Professor, Computer Science Engineering Department, WCEM, Nagpur

rahulbhandekar@gmail.com

ABSTRACT

Automated systems for identifying vehicle number plates are crucial for law enforcement and traffic management tasks. This paper introduces an innovative method that uses deep learning to recognize number plates from images and videos. The process involves detecting number plates using OpenCV's contour detection and approxPolyDP algorithm, followed by cropping and optical character recognition (OCR) using the EasyOCR library. Integration of deep learning models improves accuracy and efficiency. Through a thorough literature review, we highlight the limitations of traditional methods and emphasize the need for advanced deep-learning approaches. Our proposed method addresses these limitations by leveraging deep learning to enhance accuracy and robustness. We provide detailed implementation steps in Python and experimental results demonstrating high accuracy in number plate detection and OCR. Analyzing results, we identify strengths and weaknesses and suggest further improvement. This research contributes to advancing automated number plate recognition systems, promising improved efficiency and accuracy in various domains like law enforcement and security surveillance.

Keywords: Number plate, easyOCR, Deep Learning Model, OpenCV