Fujitsu Laboratories has been developing finger-vein and palm-vein sensors for some time (Biometric tech matches 2,048bit codes from palm vein images and Palm vein sensor accurately IDs from 10,000 hands). What's new with the company's development is slide-style palm vein authentication, which can be equipped in tablets and other handheld mobile devices.

Palm vein authentication technology utilises images captured by illuminating a palm with the safe near-infrared band light, which passes easily through the body, and reads vein patterns from the captured image. Accordingly, a palm vein authentication device's optical unit consists chiefly of an illumination component and an image capture component. In order to uniformly illuminate the entire palm, the illumination component, the widest part of the optical unit, is arranged to surround the image capture component.

As tablets and other small-scale mobile devices have become widespread, there has been interest in embedding an optical unit for vein authentication into the narrow frames of such devices, but making the optical unit smaller had been difficult.

Fujitsu Laboratories said it has succeeded in developing a compact illumination component that lights up a rectangular target area with a uniform intensity using a single LED. This was achieved by using a new compound optical element that applies the phenomenon of diffraction.

Fujitsu palmvein authentication biometrics (cr) Figure 1: Fujitsu Laboratories developed a compound optical element that utilises optical diffraction to both scatter and focus light. (Source: Fujitsu Laboratories)

The company has also developed a new verification technology that captures the complete pattern of a palm's veins, dividing the pattern into slices as the hand passes over the optical unit, which at a mere 8mm wide is able to be embedded into the frames of compact mobile devices.

Even though the optical unit's compact size results in a smaller capture area, because the hand passes over the optical unit, the pattern of palm veins is divided up for reading, enabling the entire palm vein pattern to be used for authentication. The guides make the captured area on the palm easier to reproduce. In addition, Fujitsu Laboratories developed a new algorithm for authenticating the divided-up palm vein patterns, verifying data using such features as image selection (false accept rate of 0.001%, and a one-retry false reject rate of 0.01%).

Fujitsu palmvein authentication biometrics 02 (cr) Figure 2: How the new technology is used, and how it processes information (Source: Fujitsu Laboratories)

Fujitsu Laboratories plans to continue refining its optical units and authentication algorithms with the goal of practical implementation of slide-style palm vein authentication technology during fiscal 2017.