Key Figures Handwriting Recognition Technology Based On Neural Networks

Handwritten numeral recognition is the use of computers and other automated equipment automatically identify a handwritten digital technology in our daily lives, has a very wide range of applications. In addition, the results of this research with regard also provide reference for the promotion of the character recognition, character recognition and other pattern recognition, therefore, handwritten numeral recognition has a very good application of a theoretical study of the meaning and significance.This paper studies the significance of handwriting recognition on handwritten digits BP neural network provides an overview of the basic algorithm of conventional BP neural network were introduced in handwritten numeral recognition process of the network parameters experimentally choice.According to the basic content of the neural network, and explore the use of relevant knowledge handwritten digital neural network research identified key technologies. BP neural network based on the design of the handwritten numeral recognition system combining multiple features, the system is important in three parts:pre-processing of handwritten digital images: handwritten digital image feature selection and extraction; classifier and recognition.Establish and implement a BP neural network based on handwritten digit recognition software, the classifier to identify, compare single feature four features combined into a feature vector through BP network to establish the composition of the classifier, and achieved good recognition rate.