Vehicle License Plate Localization based on Local Binary Pattern Features

Abstract

The importance of Intelligent Transportation System (ITS) is increasing because of increasing the number of vehicles on the roads. Automatic License Plate Detection (ALPD) is still a challenging task based on weather, occlusion, dirty License Plates (LP) and some other factors. In this paper Local Binary Pattern (LBP) feature is proposed for the detection of LP region. Cascaded Adaboost classifier is used for the classification of candidate region. The input images are chosen with different conditions such as occlusion, low contrast, dirty LP images and poor illumination. The proposed technique is tested on 600 images. This method achieves 96.8% detection rate.