Automatic Traffic Rules Violation Detection and Number Plate Recognition

Automatic Traffic Rules Violation Detection and Number Plate Recognition

Arun P
dept of Electronics and Communication Engineering   St.Xavier’s Catholic College of Engineering

Kanyakumari, Tamil Nadu

arunmass1603@gmail.com

Aakash S
dept of Electronics and Communication Engineering   St.Xavier’s Catholic College of Engineering

Kanyakumari, Tamil Nadu

aakashaakazh@gmail.com

 Jebin Mon M P

dept of Electronics and Communication Engineering   St.Xavier’s Catholic College of Engineering

Kanyakumari, Tamil Nadu

jebinmon1818@gmail.com

Abilesh V

dept of Electronics and Communication Engineering   St.Xavier’s Catholic College of Engineering

Kanyakumari, Tamil Nadu

jeevithajeba618@gmail.com

 Mary Little Flower B.E,M.E
dept of Electronics and Communication Engineering   St.Xavier’s Catholic College of Engineering

Kanyakumari, Tamil Nadu

          mary@sxcce.edu.in.com

Abstract—As road traffic continues to grow and traffic rule violations become more common, there's a growing need for smart systems that can automatically detect and report such incidents. This project introduces an automated solution for detecting traffic violations and recognizing license plates using deep learning, convolutional neural networks (CNNs), and optical character recognition (OCR). The system is capable of identifying common offenses such as riding without a helmet, triple riding, and running a red light, by analyzing real-time video footage. Using object detection models like YOLO, it accurately spots violators and isolates the license plate area from the video frames. Tools like EasyOCR are then used to read and extract the alphanumeric details from the plates. Once a violation is detected, the system automatically sends an email—with the offense details and a snapshot of the incident—using Python’s smtplib library. This solution is designed to support traffic enforcement agencies by making violation tracking more efficient, reliable, and automated..

Keywords—Deep Learning,CNN and OCR..