Identifying Number Plates using Feature Extraction

Abstract

A number plate can be recognized completely only when the number plate on the vehicle is neither broken nor deliberately mutilated. This pose a hinderance in recognizing these types of plates. In order to overcome this problem, one has to either compare the characters recognized from the broken plate or directly compare the features of the given plate to the database. This helps in reducing time and space complexity. Various feature extraction algorithms are available that are FAST, EIGEN, SURF, HARRIS, BRISK, MSER. Out of which SURF (speeded up robust features) is preferred. In our proposed approach, we compared features of given plate with the saved plates features. Based on the matching accuracy best matched plate is selected as output.

Keywords: image processing, number plate extraction, feature extraction, SURF, fast, eigen, Harris, brisk, MSER

Cite this Article: Saurabh Allawadhi. Identifying Number Plates Using Feature Extraction. International Journal of Image Processing and Pattern Recognition. 2019; 5(2): 20–26p.