To achieve to personal information as fast and correct a form is an important problem in official or private institutions and organizations. To disruption of the works are necessary that should be entered as quickly and without error of all numbers of card in identity recognition. In this proposed study, an optical character recognition system which recognizes as real time Republic of Turkey Identification Card Number is developed. This system consists of five stages. These stages are preprocessing, segmentation, feature extraction, classification and correction. The features of digits are created with wavelet transformations of different types and degrees from segmented digit images. Republic of Turkey Identification Card Numbers could recognize by great a success rate as real time with Artificial Neural Networks (ANN) are trained with these features.