The International Journal of Multimedia & Its Applications (IJMA) Vol.4, No.5, October 2012 DOI : 10.5121/ijma.2012.4503 39       Lama Hamandi 1 , Khaled Almustafa 2 , Rached Zantout 3 and Hasan Obeid 4 1 Electrical and Computer Engineering Dept., American University of Beirut, Beirut, Lebanon lh13@aub.edu.lb 2 College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia kalmustafa@psu.edu.sa 3 Math and Computer Science Dept., Beirut Arab University, Beirut, Lebanon rached.zantout@gmail.com 4 Zawya, Beirut, Lebanon hasanobeid@gmail.com ABSTRACT In this paper, the “character recognition” approach to recognizing a vehicle license plate is used for localizing Saudi license plates. The proposed algorithm filters out all possible objects from the license plate image and focuses on the resulting objects. The coordinates of the center point of the bounding box for these objects is calculated and then possible alignments between these objects are checked. After finding the aligned objects, the recognition algorithms are applied to differentiate the numbers from the letters in the plate. KEYWORDS License Plate Recognition, Plate Localization, Character Recognition 1. INTRODUCTION Automatic License Plate Recognition (ALPR) systems process images to find license plates and recognize the plate characters. ALPR is usually divided into two major parts, localization and recognition. Localization locates the part of the image which contains the license plate, while recognition identifies the individual characters on the assumption that the image being identified is that of a license plate. Localization algorithms can follow one of three approaches [1]. The first approach is color-based. It relies on character colors to distinguish between the characters and the background. The second approach is edge-based. It uses lines which are the edges of the characters and their boundaries. The third approach is texture-based. It locates a license plate by looking for specific patterns or textures in the image. In this paper, localization of the plate is done through the recognition of the possible characters in the plate. Recognizing these characters is done by applying three different recognition algorithms