White Paper: Sensors/Data Acquisition

Measuring Near-Infrared (NIR) Light Sources for Effective 3D Facial Recognition


Near-infrared (near-IR or NIR) light sources include LEDs and lasers, which enable quick, automated sensing applications such as facial recognition, gesture recognition, eye tracking, driver awareness monitoring, and automotive lidar. Applied in many situations where light is cast directly into a user’s face, eyes, or field of vision, NIR wavelengths are chosen for sensing because they are imperceptible to the human eye and do not cause visual disruption. NIR light sources must be tested both to ensure the light’s radiant intensity is safe for human vision, and to ensure adequate and uniform intensity across the full angular light distribution so sensing devices can send and receive light properly.

Depending on their application, NIR sources distribute light in different ways. Some sources — known as “flood” sources — emit diffused light to gauge object presence and proximity. Common in facial recognition, diffractive optical elements (DOE) are applied with NIR laser sources to distribute the light in “structured light patterns” (i.e., dots or a grid). Deformations between the grid cast on a user’s face and the returning light received by a device sensor are measured to interpret a user’s 3D facial features, enabling biometric device security. These dot patterns must be distributed properly, in the correct patterns, with safe and adequate intensity for device operation.

In this paper, learn how NIR light is used to enable 3D sensing in facial recognition systems. This paper explains the importance of measuring the radiant intensity of emissions from NIR light sources like LEDs and lasers, and the challenges associated with using traditional measurement methods to obtain complete data over a full angular distribution. An integrated NIR Intensity Lens solution is described with advantages for evaluating NIR emitters to ensure the quality of facial recognition systems in devices such as smartphones, laptops, and automobiles.

Don't have an account? Sign up here.