Iris marking normalization process method

Abstract

A normalized processing method for the iris texture includes such steps as using Log operator to extract the margin of sampled iris image, using Hough transform to locate the center and radiuses of internal and external circles of iris, using modules to correct physiological deviation and sampling deviation, correcting the non-uniform zoom, and developing the sample texture points to form a rectangular iris texture image.

Description

technical field [0001] The invention belongs to the technical field of eye iris recognition, and in particular relates to an iris texture normalization processing method. Background technique [0002] Identity recognition has a wide range of needs in modern high-speed information-based social life. Biological characteristics closely related to human beings, such as appearance, fingerprints, irises, voice, gait and signature, etc., all have the characteristics of "everyone is different, they will remain the same throughout their lives, and they will be carried with them". Among them, the iris is particularly "unreformable and difficult to disguise", and will become an important way for future identification. [0003] The iris is located in the ring-shaped part between the black pupil and the white sclera of the human eye, which presents a radial structure from the inside to the outside, including many interlaced subtle features similar to spots, fine lines, crowns, crypts, e...

AI Technical Summary

Problems solved by technology

However, when there is a relatively large deviation between the inner and outer centers of the iris (when the iris image resolution: 350*220, the deviation between the inner and outer centers: more than 3 pixels) or when the pupil is dilated (that is, a non-standard iris), if the traditional normalization is still used Algorithm, which will cause the loss of part of the texture of the iris edge or mistake the pupil and sclera area for the iris area

Such problems will seriously affect the accuracy of iris coding, and even lead to recognition errors

[0015] This is because the traditional iris normalization algorithm ignores two important problems: (1) the eyeball rotation during the acquisition process or innate physiological reasons cause the center of the inner and outer circles of the iris to not coincide; (2) the center of the iris caused by light stimulation Pupil constriction and dilation are not perfectly uniform

Second, during the acquisition process, the deviation of the center of the circle due to the rotation of the eyeballs

Due to the "barrel distortion effect" of macro imaging, when the eyeball turns a small angle, the iris image will be greatly distorted, resulting in a deviation of the center of the inner and outer circles

Images