Vehicle Number Plate Detection and Recognition

Vehicle Number Plate Detection and Recognition

Ankit Kaushal, Narinder Yadav, Nikhil Kumar, Sachin Yadav, Ishdeep Singla, Dipanshu Garg

Computer Science and Engineering Chandigarh University

Punjab, India

Abstract-

Real-time traffic monitoring systems face a host of challenges, especially the conventional methods using expensive labour-intensive and time-consuming manual operations . These traditional techniques generally involve human operators and hence are bound by scalability and efficiency. However, with advances in surveillance technology and computer vision, new prospects became possible . Through the incorporated video feeds coming from security cameras and traffic monitoring systems, detection of traffic flow, counting, and analyzing the behavioural patterns of the numbered plates can be realized in modern approaches.

Another striking project in the vision is the project of traffic camera analysis, which suggests an entire system for real- time analysis of cameras regarding the detection of traffic. As the system will automate all  processes, it is envisioned to give both practical and  educative insights into the issue of traffic: vehicle detection and number plate recognition. The system does this by combining enhanced methods of image processing with object detection and tracking, enabling the system to closely observe the flow of traffic. All data analyzed via this system is real-time, meaning that each    count-and    recognition    of    a     car-is     valid. The implementation will use the Python language, and hence it has very potent libraries for image processing as well as machine learning. The presented work will thus equally look at the technical challenge about traffic monitoring in real-time by further discussing the issues in terms of scalability related to practical applications in infrastructural smart cities.

Keywords: Real-Time Traffic Monitoring, Automatic Number Plate Recognition, EasyOCR, OpenCV.