Method of textured contact lens detection
Inventors
Czajka, Adam • Bowyer, Kevin • Fang, Zhaoyuan
Assignees
Publication Number
US-11768926-B2
Publication Date
2023-09-26
Expiration Date
2039-10-24
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Systems and methods for detecting textured contact lenses in an iris image are disclosed. Initially, “K” images are taken of an eye in near-infrared light, each for a different positioning of the illuminant. Three-dimensional properties of the hypothetical iris are estimated based on image processing methods. The variability of the estimated three-dimensional properties is calculated, and a classification into two classes is made based on the variability. A low variability denotes that an authentic iris pattern was presented to the sensor, whereas a high variability denotes that the sensor observes an eye wearing a textured contact lens. The systems and methods disclosed allow for detecting presentation attacks to iris recognition systems/sensors based on presentation of an eye wearing a textured contact lens in an automatic and accurate way.
Core Innovation
The disclosed invention provides systems and methods for detecting textured contact lenses in iris images by leveraging differences in three-dimensional surface properties between an authentic iris and an iris with a textured contact lens. The method involves capturing multiple images of an eye with near-infrared light from different angles using at least two illuminants positioned relative to a camera. By examining how shadows and surface irregularities appear in these images, the system can estimate the three-dimensional structure of the observed iris region.
An authentic iris is generally flat, so the surface normal vectors calculated from images under different illuminations will be similar, demonstrating low variability. In contrast, a textured contact lens introduces irregularities, causing greater variability in the estimated normal vectors. The variance of these normal vectors, especially in the nonoccluded iris annulus, serves as a key indicator. If the variance exceeds a predetermined threshold, the system classifies the presentation as involving a textured contact lens.
This capability addresses the critical problem where attackers can bypass iris recognition systems by wearing textured contact lenses that simulate a target iris or obscure the attacker's own iris. Prior techniques either required additional complex hardware or failed to distinguish textured contact lenses from authentic irises. The disclosed approach can be implemented with existing sensor hardware that features multiple illuminators, enabling accurate and automatic detection of textured contact lenses in iris recognition scenarios.
Claims Coverage
The patent includes two independent claims focusing on a method and a computer system for detecting textured contact lenses using surface shape analysis of iris images.
Computer-implemented method for detecting textured contact lenses using multi-angle illumination and surface variance analysis
This method involves: - Illuminating the eye with at least two illuminants placed at different angles relative to a camera. - Acquiring iris images using the camera, with each image taken under illumination by an individual illuminant. - Determining the presence of a textured contact lens by: 1. Calculating normal vectors to the observed surface from the iris images. 2. Identifying the nonoccluded portion of the iris annulus. 3. Determining the variance of the normal vectors within the nonoccluded iris region. 4. Comparing the variance to an acceptance threshold and outputting a result indicating the presence of a textured contact lens if the variance exceeds the threshold.
Computer system for detecting textured contact lenses using analysis of estimated surface shape from iris images
This system comprises: - One or more processors and storage devices executing computer-executable instructions. - Receiving multiple iris images, each acquired during illumination with one of at least two illuminants at different camera-relative angles. - Determining the presence of a textured contact lens by: 1. Calculating surface normal vectors based on the iris images. 2. Calculating a nonoccluded portion of the iris annulus. 3. Determining the variance of the surface normal vectors within the nonoccluded region. 4. Comparing the variance to an acceptance threshold and outputting a result indicating the presence of a textured contact lens if the variance exceeds the threshold.
The inventive features center on the use of multi-angle image acquisition, photometric surface normal calculation, selective iris region analysis, and automated variance threshold-based classification to distinguish textured contact lenses from authentic irises in iris recognition systems.
Stated Advantages
The invention enables accurate detection of textured contact lenses in iris recognition systems, addressing a critical vulnerability to presentation attacks.
The methods can be implemented with existing commercial iris sensor hardware, often requiring no hardware updates.
The approach allows for automatic and reliable classification of genuine eyes versus eyes with textured contact lenses using a computational method.
It can be used with sensors compliant with ISO/IEC 19794-6, leveraging standard near-infrared illuminators already present in most devices.
Documented Applications
Presentation attack detection (PAD) in iris recognition sensors to prevent unauthorized access using textured contact lenses.
Interested in licensing this patent?