Automatic License Plate Recognition (ALPR) for Bangladeshi Vehicles
This paper presents Automatic License Plate extraction, character segmentation and recognition method for license plate of Bangladeshi vehicles with chain code and neural network. In Bangladesh, license plate models are not followed strictly. Characters on plate are in Bangla and English languages and also are in one or two lines. Due to dissimilarity in the model of license plates, vehicle license plate extraction, character segmentation and recognition are key issue. Different types of algorithm already applied and the performance is examined for English license plate. We describe the license plate extraction, character segmentation and recognition work, with Bangla characters. License plate extraction is performed using Sobel filter, connected component analysis and morphological operations. Character segmentation is performed in different levels by using scanning the binary image horizontally and vertically and connected component analysis. Character recognition is carried out using chain code generation and stored knowledge of the network.
