Performance analysis of localization algorithms applied on real time license plates of different countries

Abstract

A very large number of localization algorithms have been used till under various domains. But the subject of our research paper deals with the domain of VPR i.e. vehicle plate recognition system. The accuracy of the license plate recognition system depends on the performance of the localization algorithm. It is a computationally intensive process which takes a lot of time to localize the license plate. In our research paper we will discuss different types of localization algorithms and which of them should be used for a particular application. Different countries have their own types of license plates for instance some use single line horizontal license plates while others use multi line non horizontal and differently located number plates. Some of the widely used localizations algorithms which are used in the neural network recognizer are global threshold scheme, NiBlack's threshold.