基于统计特征和环数的车牌字符识别算法研究

o solve the deficiency of feature extraction method in vehicle license plate character recognition, a new method based on ring number feature and multi-statistical feature extraction is presented. First,the 34 characters of the license plate are divided into 3 categories according to the number of rings:no ring, a ring and two rings. For each character class, after pre-processing,different kinds of features are extracted，including stroke statistical feature on horizontal and vertical , inflection point statistical characteristics of four outsides of characters, outline of statistical features of four outsides of characters and pixel statistical characteristics of characters.These features are then converted into a lower-dimension feature vector set,and then test the features extraction using the improved BP neural network.The experiment results that the recognition of license plate characters based on statistical characteristics and ring number recognition method is effective to improve the similar characters recognition and robust.