Automatic Number Plate Recognition (ANPR): an Experiment

EasyChair Preprint 7834

Abstract

The objective of ANPR is to extract vehicle license plate information from number plate of vehicles. As the traffic control and vehicle owner identification is a major issue in every country, it is important to develop such device which automatically detect those vehicle owner who violates traffic rules and drives fast. There are many ANPR systems are present but there is challenging factor like accuracy of extraction, speed of vehicles, lightening condition, quality of images. In this paper, different methods of ANPR and emerging technologies are used to get accurate result. The important work is the detection and recognition of the number plate which is accomplished by the Convolution Neural Network (CNN). Reason to choose CNN is the high accuracy of 90% even with very low training size. We categorize many ANPR techniques as per their features they used in each stage and compare them in terms of their advantages and disadvantages, accuracy and processing speed.

Keyphrases: ANPR Deep, learning image, recognition

@booklet{EasyChair:7834,
  author    = {Manul Gupta and Arpit Agarwal},
  title     = {Automatic Number Plate Recognition (ANPR): an Experiment},
  howpublished = {EasyChair Preprint 7834},
  year      = {EasyChair, 2022}}