Automatic Number Plate Detection in Vehicles using Faster R-CNN

Abstract

The paper is aimed to identify the number plate in the vehicles during difficult situations like distorted, high/low light and dusty situations. The paper proposes the use of the Faster R-CNN to detect the number plate in the vehicle from the surveillance camera which is placed on the traffic areas etc. The created system is used to capture the video of the vehicle and then detect the number plate from the video using frame segmentation and image interpolation for better results. From the resulted image using the technique called optical character recognition is applied on that image for number recognition. These number are given as input to the database to retrieve data like vehicle's name, owner name, address, owner mobile number, etc. The performance of this system is measured using in a graph model. The proposed system is able to achieve a 99.1% accuracy to detect the number plate of the vehicle and show the vehicle's owner information.