METHOD OF RECOGNIZING HUMAN IRIS USING DAUBECHIES WAVELET TRANSFORM

摘要

PURPOSE: An iris recognition method is provided to obtain an eye image by using a halogen lamp illumination, to preprocess the obtained eye image for extracting an iris area from the eye image, and to identify a person by using the preprocessed image. CONSTITUTION: The method comprises steps of obtaining an eye image of a person by using an image acquisition equipment(200), extracting an iris area from the eye image via a preprocess step, and converting the iris area data in a polar coordinate system to extract an iris pattern(210), performing the Daubechies wavelet conversion operation on the converted iris pattern and extracting characteristics on the iris(220), generating a binary characteristic vector by applying a K level quantization function on the extracted characteristics(230), measuring a similarity by comparing the generated characteristic vector with stored data on the person, and determining an identification of the person according to the measured similarity(240). The Daubechies wavelet conversion step(220) includes steps of multiply dividing the extracted iris area, extracting a high frequency component area from the divided iris area along the x and the y axis, increasing a division repetition number of the iris image to a designated number, storing the high frequency component area as characteristic data, extracting a low frequency component area from the divided iris area along the x and the y axis, and offering the low frequency component area in order to repeatedly apply the Daubechies wavelet conversion to the low frequency component area.