A template-based model for license plate recognition

Using video camera to manage the cars is gradually adopted in many lands of applications, such as electric payment in the tailgate and car parking management. An automatic recognition model of automobile's license plate number is proposed in this paper. The designed system is expected to have high recognition accuracy and reliability such that the goal of automatic recognition can be achieved. We present a system to recognize the license number in the acquired image captured from a video camera. The recognition process of our system contains four major steps. First, the system tries to locate the probable position of the license plate within the acquired image by using gradient analysis and image processing. Second, our model estimates the image parameters needed to normalize the license plate and uses the cross-correlation to detect the skew of the license plate and rectify the tilt. Third, we use a template technique to recognize the characters in the license plate. Finally, we use the information gained from the previous step to analyze the probable license numbers. We illustrate the designing processes and give the experimental results from the proposed model. Based on the experimental results, the proposed system can effectively recognize the license number. The time needed to recognize a license plate takes only 1.5 seconds.