Automatic vehicle identfication for Argentinean license plates using intelligent template matching

Abstract

The problem of automatic number plate recognition (ANPR) has been studied from different aspects since the early 90s. Efficient approaches have been recently developed, particularly based on the features of the license plate representation used in different countries. This paper focuses on a novel approach to solving the ANPR problem for Argentinean license plates, called Intelligent Template Matching (ITM). We compare the performance obtained with other competitive approaches to robust pattern recognition (such as artificial neural networks), showing the advantages both in classification accuracy and training time. The approach can also be easily extended to other license plate representation systems different from the one used in Argentina.