Research of ID Card Recognition Algorithm Based on Neural Network Pattern Recognition

摘要

How to quickly and correctly identify the second generation ID card code has important practical significance in real life. The algorithms have been presented about the recognition and verification code of ID cards. In order to let computer own the knowledge of code of character image, firstly, transform the image of ID cards to binary image, then segment image character of ID card code and display image character using image processing technique, secondly manual input corresponding correct code of character image, thirdly character image is classification stored in the corresponding folder according to the corresponding correct code. These character images are training samples, which are trained using the matlab toolbox of the neural network pattern to generate training model. Using the generated the training model to identify code of the ID cards. The validation algorithm is based on the generated the training model to identify code of ID cards. Compare the consistent of code of ID cards to identify the code, if recognition accuracy rate is not high, the training model need to be trained again, until the recognition accuracy rate is more than 95%. Neural network pattern recognition and verification algorithm can take advantage of the training model to perfectly realize identification and validation the code of the second-generation ID card, the algorithms have application value in the real work.