Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-9665772-B2
Publication Date
2017-05-30
Expiration Date
Abstract
A method and apparatus for biometric iris matching comprising pre-processing an input image capturing one or more eyes to produce one or more rectified iris images, coding the one or more rectified iris images into one or more multiresolution iris codes and matching the one or more multiresolution iris code with a set of stored multiresolution iris codes to determine whether a match exists.
Core Innovation
A biometric iris matching computer system receives at least one input image of an eye and generates one or more edge maps corresponding to each of the at least one input image. The system segments a pupil area in each of the at least one input image and segments an iris area from a sclera area in each of the at least one input image. The system synthesizes an artificial pupil on the at least one input image within the segmented pupil area.
The system produces rectified iris images and transforms the rectified iris images into transformed iris images that are invariant coordinate system image representations. The system performs normalization in an invariant coordinate system, generates multiresolution iris codes from the transformed iris images, and masks specular reflections, eyelid/eyelash regions, and unstable local phase regions. The coding converts the rectified iris into a polar iris image and generates multiresolution iris codes using bandpass filters and Gabor filters with masking of specular/eyelid/unstable regions.
For matching, the system aligns iris codes using constrained barrel shift and performs smooth flow estimation to reduce residual distortion. The system then computes match scores, including Hamming distance, and also supports alternative match measures including phase difference histogram, mutual information, and conditional entropy. The system is implemented as a computer system including a processor and memory storing instructions to perform the receive, generate, segment, transform/correct, code, and matching operations.
The document further describes image preprocessing that corrects for tilt/obliquity and corrects corneal distortion using an estimated angle of tilt with respect to a camera. It includes warping actions such as stretching in the estimated direction of tilt and applying affine and/or projective transformations, including aligning iris and pupil contours into concentric circles. It also describes selecting a pupil contour based on best fitting circle and best inscribed circle for pupil segmentation, and selecting the pupil contour from candidate pupil contours constructed from edge maps.
Claims Coverage
The provided independent claims cover three main groups of inventive features: (1) a computer system pipeline that segments pupil and iris and synthesizes an artificial pupil, (2) tilt estimation and corneal distortion correction, and (3) pupil contour selection and iris segmentation from sclera via a best circle-based contour selection approach. Together, the independent claims define core computer-system and method-level steps for biometric iris matching, including edge maps, pupil/iris segmentation, geometric correction, and representation/selection for subsequent matching.
Edge map generation, pupil and iris segmentation, and artificial pupil synthesis
Receive at least one input image of an eye, generate one or more edge maps corresponding to each of the at least one input image, segment a pupil area in each of the at least one input image, segment an iris area from a sclera area in each of the at least one input image, and synthesize an artificial pupil on the at least one input image within the segmented pupil area.
Angle of tilt estimation and corneal distortion correction
Receive at least one input image of an eye, estimate an angle of tilt of the eye in the at least one input image with respect to a camera that captured the image, and correct corneal distortion based on the estimated angle of tilt.
Best circle-based pupil contour selection and sclera-based iris segmentation
Receive at least one input image of an eye, segmenting a pupil area in an eye detected in at least one input image, selecting a pupil contour based on one or more of a best fitting circle and a best inscribed circle, and segmenting an iris area from a sclera area in each of the one or more eyes in the input image.
Across the independent claims, the inventive coverage centers on (i) edge-map-driven pupil and iris segmentation with artificial pupil synthesis, (ii) estimating an eye tilt angle relative to the camera and correcting corneal distortion based on that estimate, and (iii) selecting a pupil contour using best fitting and best inscribed circles while segmenting the iris area from a sclera area.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Authentication.
Device authorization.
Security.
Medical access.
Marketing/customer tracking.
Interested in licensing this patent?