Iris Edge Positioning Method and Iris Recognition System

Abstract

The iris positioning method of the present invention locates the inner boundary and the outer boundary of the iris according to the iris image, and when the inner boundary is located, a series of circles are selected in the iris image, and the average pixel of the rectangle with the largest area in its inscribed rectangle is calculated for each circle, Calculate the respective average pixels of a plurality of predetermined small rectangles around the circle, and use the circle of a small rectangle whose average pixel of the small rectangle is greater than the average pixel of the large rectangle as a candidate for the inner circle boundary, and calculate the The respective gradients of the selected multiple sampling points and the cross-correlation coefficients between gradient vectors and vectors are selected, and the circle with the largest sum of cross-correlation coefficients is selected from the inner circle boundary candidates as the inner boundary; when locating the outer boundary, select in the iris image A series of circles that have been determined as inner circles are used as candidates for the outer circle boundary, and the cross-correlation coefficient is calculated according to the variance of the predetermined area inside and outside the circle calculated for each sampling point on each circle, and the cross-correlation coefficient with the largest sum is selected circle as the outer boundary.

Description

technical field [0001] The invention relates to an iris edge positioning method. Background technique [0002] Iris recognition is a biometric identification method that uses the iris to identify people. The iris has a wealth of texture information, and the iris recognition system is the most accurate among biometric systems (such as other fingerprints, faces, voiceprints, palmprints, etc.). Because of these advantages, iris recognition has begun to be widely used. In general, an iris recognition system includes iris positioning, normalization, feature extraction and comparison. [0003] In iris localization, the inner and outer boundaries of the iris and the upper and lower boundaries of the eyelids are first located, and then the reflective areas and eyelashes are detected and removed. In the normalization step, Cartesian coordinates are converted to polar coordinates, and the localized iris area is converted to a fixed-size rectangular area. Since the normalization ste...

AI Technical Summary

Problems solved by technology

This method cannot find the correct iris edge when the iris edge is noisy such as there is a reflection area

Images