Multistage morphology-based license-plate location algorithm

Abstract

This paper describes a new license-plate location algorithm using a sequential multistage approach, where each stage consists of a different sub-algorithm. All the proposed subalgorithms make use of morphological operations performed over both gray-scale and binary images. In order to ascertain the performance of the proposed algorithm, and to fairly compare it with license-plate location algorithms developed by other authors, a well-defined evaluation method is used. Results have shown that the proposed algorithm produces at least one accurate license-plate location hypothesis for 97.18% of 673 images acquired at a highway toll fence.