Paper Key : IRJ************759
Author: Omkar Kakasaheb Rajale,Sachin Kumar Yadav,Rishabh Choudhary,Sumaira A. Shaikh
Date Published: 01 Nov 2022
Abstract
Identity of license plate inclusive of characters and numerals is proposed using CNN. This is executed by way of amassing the numerals from (zero-nine) and characters from (A-Z) from which a dataset is constructed to train convolutional neural networks (CNN). The challenge is quite hard because of non uniform illumination situations the variety in plate codecs. License plate recognition is the effective manner for figuring out routinely , there also exists different algorithms but they not as effective in turning in pleasant effects. Number Plate recognition algorithms in pictures and videos typically includes 3 stages 1)license plate extraction 2)character segmentation 3)character by character recognition. With the skilled model, character recognition device is designed and implemented. In this thesis we make use of object detection algorithm YOLOV3 with the help of network structure Darknet.