Description: Vehicle license in the automatic identification system, the result of natural factors or sampling factors make the original rules of the print characters have a distortion, to the character recognition has led to great difficulties. In this paper, feature extraction, based on the BP network used for classification, and additional linear perceptron to realize the effective recognition of words. The method is simple algorithm, high recognition rate, applicable to a wide range of high-noise environments print character recognition.
File list (Check if you may need any files):
Download