Abstract
Many areas include public transportation, hospitals and shopping mall use the computer technologies for security reasons. The need for integrating the technology with human needs is increased dramatically. Car license plate recognition is one of those technologies that help people to secure themselves from different attacks either attacks that affect their lives or their properties. There is an increasing demand for a car plate recognition system that can recognize the character in the Saudi plate specifically. Many challenges found in the Saudi license plate which make this a difficult task. Unlike other countries, Saudi plate has Arabic letters and numbers, English letters and numbers, horizontal and vertical lines and finally the logo and the abbreviation of Kingdom of Saudi Arabia. The aim of this paper is to design a system that could recognize the plate number efficiently. The bi-linear interpolation algorithm was used to resize the license plate image to have a fixed size of all license plates. After resizing the image, different preprocessing techniques were applied to remove the unnecessary parts. Finally, an Optical Character Recognition (OCR) that is based on Artificial Neural Network (ANN) classifier is used to recognize the characters. Matlab tool was used for designing and testing. Several experiments conducted to calculate the accuracy and show the efficiency of our design. The results show that our system has 92% accuracy and can recognize both Arabic and English letters and numbers.