Vehicle identification systems for Turkish private license plate standarts

In the global era, experienced revolution in production processes thanks to the development of new technologies, population growth, and intensified urbanization results in increase in the number of vehicles. This situation has generated world wide interest in and demand for Automatic Vehicle Identification Systems (AVIs). This thesis introduces a new system designed to read the license plates of private vehicles based on the Turkish Licence Plate Numbering System. In the literature review part, historical development of the methods has been utilized for reading the licence plates on vehicles were examined. The main part of the study explains the procedures and techniques used at the image acquisition and processing and character recognition steps to design the new system. Image processing algorithms, inductive learning and template matching techniques were utilized in this study. More specifically, for license plate recognition, following techniques were integrated through morphological filtering: resizing the captured image, gray scale digital image, histogram equalization, Thresholding and Smearing algorithms. In character recognition and license plate identification step, morphological filterings, template matching and inductive learning were used. Experiments were conducted in different conditions and with the images from the front and back sides of the vehicles and promising results were attained in terms of the performance and the accuracy of the system.