Abstract - Despite the abundance of commercial and academic techniques for Automatic Number Plate Recognition (ANPR) [1], the majority of current algorithms concentrate on a particular region of the number plate (NP) and frequently examine data sets with roughly frontal images. To identify the car, we use a modified version of YOLO, and after that, we localize the license plate. Additionally, we enhance an optical character recognition network (OCR-Net) [1] to recognize the numbers and letters on license plate [2] by CNN techniques. Our approach works well with various vehicle kinds. Images captured by a variety of cameras show vehicle license plates from various nations, automobiles at various distances, and more.
Keywords: Number Plate, Convolutional Neural Networks, Optical Character Recognition.
|
DOI:
10.17148/IJARCCE.2023.125216
