A license plate segmentation algorithm based on MSER and template matching

Intelligent Transportation Systems (ITS) is becoming more and more popular in daily life. The License Plate Recognition (LPR) is an important part of the ITS, and it is also a basic part in traffic management. Generally speaking, the LPR system consists of three parts: license plate location, license plate character segmentation and character recognition. In this paper, a license plate character segmentation algorithm based on Maximally Stable Extremal Region (MSER) and template matching is proposed. The MSER detector is used to detect the candidate character regions and the template matching is in order to accurately find the location of the seven license plate characters. The algorithm is tested on a dataset which is achieved through the license plate location. The dataset includes two categories of license plate: one-row plate and two-row plate. The average accuracy of this algorithm is 96.08%.