Advanced sensing capabilities have been used to enable biometric access solutions in the past. Most prominent are biometric face scans, iris scans, or fingerprint- / hand-print scans. Emerging recently are solutions around unique cardiac patterns or palm vein scans are being evaluated. Let’s take a closer look to the state-of-the art solutions first.
Biometric face scans typically capture the following features which are then compared for size, thickness, placement, shape or contour: cheek bones, eyes, nose, mouth, lips, chin and jaw line, forehead and scalp, ears. Critical infrastructure requires 100% biometric match whereas for most commercial systems a 80-90% match is sufficient. Protection against spoofing or faking (e.g. putting a photo in front of the scanner instead of real person) require true 3D face data capturing. For doing so, either dual-camera stereovision or single camera structured light scanning (where a single camera measures the shift of defined dot pattern projection) are being used.
Another biometric scanning approach uses iris scans. Each person has a unique iris in texture, size and color. A combination of near-infrared or visible light illuminators plus a digital camera are at the core of biometric eye scanning system. Typical distances between iris and scanning system are ~8-40cm. Iris scanning is among the most secure scanning technologies as it uses up to 240 reference points for deciding on a match (in comparison finger scans typically rely on ~60 reference points for a match).
Finally, finger-print- / hand-print readers scan the tiny friction ridges on the ends of fingers and thumbs, which are unique for each person. The scan can be performed optically (fast but sensitive to dirt), capacitively (fast but sensitive to wet or dirty hands) or using ultrasonic (even 3D scan, but slow). Irrespective of the scanning method, a map of the friction ridges is created which is evaluated during access scanning. Optical scans are typically about 512x512 pixel for a ~2.5cm2 image at ~500dpi and 256 grey shade levels. Live finger detection algorithms are deployed to detect fake fingers which the follow the same methods as known for vital sign monitoring: heart rate, blood oxygen etc.
For all of the different scanning approaches, including vital sign monitoring, ams OSRAM provides a broad offering of illumination solutions, ranging from flat -side- looking LED devices, extremely thin and low- power backlighting LEDs to new-infrared laser solutions for flood illumination and dot pattern projection. On the sensing side, different imager options, ranging from 0.5 to 2.2Mpixel resolutions, are available; some include integrated low-power wake-up modes for best possible energy saving. If implementing full 3D sensing is too complex, anti-spoofing can also be implemented using ams OSRAM’s latest generation of multi-zone time-of-flight proximity sensors which are suitable for efficient presence detection methods.