This week, the European Patent Office published two patent applications from Apple relates to Optic ID, Apple's form of biometric identification used for their upcoming Apple Vision Pro Spatial Computer. Below is the video segment of the WWDC23 keynote's introduction to Optic ID.

 

  

In Apple's European patent background they note that an eye or gaze tracker is a device for estimating eye positions and eye movement. Eye tracking systems have been used in research on the visual system, in psychology, psycholinguistics, marketing, and as input devices for human-computer interaction. In the latter application, typically the intersection of a person’s point of gaze with a desktop monitor is considered.

Biometric authentication technology uses one or more features of a person to identify that person, for example for secure, authenticated access to devices, systems, or rooms. In a typical scenario, in a registration process one or more images are captured of the features being tracked (e.g., images of a person’s iris(es)), and the images are processed to generate a set or vector of metrics that are unique to, and thus uniquely identify, that person.

When the person attempts to access the device, system, or room, images of the person’s features are again captured and processed using a similar algorithm to the one used during registration. The extracted metrics are compared to the baseline metrics and, if the match is sufficiently good, the person is allowed access.

Apple's patent covers embodiments of methods and apparatus for biometric authentication are described in which two or more cameras are used to capture images of biometric features or aspects for analysis to identify and authenticate a person.

In embodiments, an imaging system includes at least two cameras that are used to capture images of a person’s iris, eye, periorbital region, and/or other regions of the person’s face, and one or more features from the captured images are analyzed to identify and authenticate the person (or to detect attempts to spoof the biometric authentication).

Embodiments may, for example, be used to capture images of the user’s iris using two or more eye cameras for biometric authentication. In some embodiments, instead of or in addition to the iris, two or more cameras may be used to capture biometric aspects or features of the eye, periorbital region, or user’s face including but not limited to the eye surface, eye veins, eyelid, eye brows, skin, or nose, and use the biometrics alone or in combination to perform biometric authentication. In some embodiments, feature sizes and geometric relations to other features may be included as biometric aspects.

Embodiments of biometric systems or algorithms may use images from at least one of the two or more cameras (two or more per eye, in some systems) that capture images from different viewpoints of the user’s eye, periorbital region, or face to perform biometric authentication. Having at least two cameras per eye may, for example, allow the imaging system to capture images of the eye region from different angles, and allow for switching to a more favorable point of view (pose as location and orientation), and in some embodiments may allow for two or more images captured by two or more cameras to be combined for use in biometric authentication.

In some embodiments, an algorithm executing on a controller coupled to the two more cameras may dynamically determine which image(s) captured by the two or more cameras are to be used for biometric authentication, for example using one or more objective criteria to evaluate the quality of the captured images. The objective criteria may include one or more of, but are not limited to, exposure, contrast, shadows, edges, undesirable streaks, occluding objects, sharpness, uniformity of illumination, absence of undesired reflections, etc. In addition, properties of the region being captured by a camera may be evaluated to determine quality, for example an overlap of a part of the eye by an eyelid may obscure at least part of a feature in an image captured by one camera while the feature is more visible in an image captured by a second camera.

In some embodiments, an algorithm executing on a controller coupled to the two more cameras may combine information from two or more images of an eye, the periorbital region, or portions of the face captured by at least two cameras to process aspects and features extracted from the combined images.

Apple's patent FIGS. 5A and 5B below illustrate a biometric authentication system that combines different biometric aspects including a camera system used to identify the user's Iris; FIG. 7 illustrates a biometric authentication system that uses multiple cameras; FIGS. 12A illustrates a system that includes light sources that emit light at multiple wavelengths to sequentially capture images at multiple wavelengths.

Apple's patent FIG. 20 above illustrates an example head-mounted device (HMD) that may include components to capture Iris characteristics.

Embodiments may provide methods that improve the performance and robustness of an imaging system, and that make the imaging system adaptable to specific users, conditions, and setup for biometric authentication using the eyes and periorbital region, gaze tracking, and anti-spoofing.

While, conventional eye tracking systems focus on specular reflections or glints for gaze tracking, embodiments may focus on other aspects such as providing uniform, good contrast on the iris or other regions of interest, reducing or illuminating shadows on regions of interest, and other improvements for biometric authentication applications.

For finer details, review Apple's patent application EP4217919 published in Europe on August 02, 2023. Another patent filed in Europe associated with the '919 patent is EP4217920 titled "Automatic Selection of Biometric based on Quality of Acquired Image.