Research On License Plate Character Recognition Key Technologies

With the development of national economy, the amount of vehicles has increased rapidly. Vehicles make in deed our life convenient, but also cause a lot of problems. Therefore, to adopt Intelligent Transportation System (ITS) has been more and more important. In addition, license plate automatic recognition technique is the kernel of ITS. License plate recognition technology has an important academic value will also has an enormous social and economic benefits. This paper studies some basic license plate recognition algorithm and recognition technology. The study includes the following three aspects:1. In stage of character image pre-processing, studies a character image enhancement, including histogram equalization and median filter. Studies the character image binarization deeply. Compares three types of license plate character image interpolation method, and find the most suitable interpolation method for normalized.2. In the feature extraction stage, studies the outline of each character feature, texture feature, interior feature and skeleton feature. Analyzes the Gabor filter parameter selection deeply. Apply the three kinds of feature extraction method to feature extraction of license plate characters. Obtains satisfactory recognition results.3. In the identification stage, Use the matching algorithm. Analyzes the Shortcoming of the BP neural network in character recognition. Use the support vector machine (SVM) for character recognition. And adopt a method of calculating reliability for recognition results. Combines the Several of Character features and Recognition effectively base on the credibility. Recognition results are satisfactory.