Character Recognition System Based on Back-Propagation Neural Network

Abstract

According to the characteristics of vehicle license plate, recognition algorithm was proposed based on back-propagation (BP) neural network. Classifier was divided into Chinese characters classifier, English letters classifier, English letters and numbers mixed classifier, and digital classifier these four kinds of classifier in the algorithm. This neural network design can effectively simplify the network structure, improved recognition accuracy and speed. BP algorithm went along improvement as the defects of the standard BP algorithm which had slow convergence and easy to fall into local minimum points. Through simulation experiments, the character recognition system not only has a higher recognition rate, but also has better neural network robustness to decrease failures, that is having good robustness characteristics.