Deep Learning based Automatic Vehicle Number Plate Recognition for Vehicle Authentication and Intelligent Security System

Vehicle number plate recognition is useful in tollway, parking area, highways, traffic signals, etc. In all these places vehicle number needs to be registered either manually or systematically for future reference. By registering the vehicle number, officials can easily trace the vehicle owner’s details from the database. So we can identify the cause for the problems like robbery, kidnap, and illegal activities. In the proposed work, vehicle number plate recognition has been automatically collected and maintain the database of the vehicle number with the help of deep learning algorithm. The system proceeds with the following steps. Firstly, vehicles need to be identified from CCTV. For this process, a coherent model namely Faster R-CNN is used. Secondly, number plates from the identified vehicles need to be separated to recognize the numbers. Finally, OCR is used in the extracted image to spot the numeric and alphabets in the number plate. The accuracy of 98% achieved in this system.