Ruled-based and stroke composition technique for efficient license plate recognition

In this paper, we address the issue of recognizing standard car license plate. We propose to have only two stages of recognition process that are localization and recognition. For localization, instead of searching the region for the plate, we directly locate the alphanumeric characters of the car plate to remove issues such as plate size or shape variations, broken plate as well as plates on black colored vehicles. We use ruled-based technique to locate and detect the characters of the car plates. In the second stage, we propose stroke composition technique to recognize the characters. The proposed technique first extract the stroke of the character and then stroke composition technique is used for classification. Our main goal is to avoid using a lot of training samples to handle various font sizes. The advantage of this method is that the system requires no training. From the experiment performed, the method has a correct recognition accuracy of 95%.