An Efficient Features - Based License Plate Localization Method

Abstract

This paper presents a feature-based license plate localization algorithm that copes with multi-object problem in different image capturing conditions. The proposed algorithm is robust against illumination, shadow, scale, rotation, and weather condition. It extracts license plate candidates using edge statistics and morphological operations and removes the incorrect candidates according to the determined features of license plates. We have formed a rather complete database of 269 images in different conditions. The proposed algorithm successfully detecteds the accurate location of the license plates in 96.5% cases, which outperforms the other available approaches in the literature.