Near-Infrared Sensing

Facial & Gesture Recognition

Near-infrared (NIR) light covers wavelengths of the electromagnetic spectrum from approximately 750 to 2500 nanometers (nm), which is beyond human visual perception. NIR-based sensing  provides machines with information about objects in the physical environment. When NIR light is emitted and reflected off of an object, an NIR sensor receives the reflected light or light pattern to gauge the distance, size, location, and identifying features of objects in the three-dimensional world.

NIR light is ideal for 3D sensing applications because it is invisible to the human eye, making it unnoticeable when cast into a user’s face, eyes, or environment. Proper filtering can remove the visible light spectrum for NIR sensing devices, increasing reliability for sensing in ambient light conditions or poor visibility conditions like rain or fog. However, because the human eye does not respond to NIR light, it is important to measure the emissions from NIR LEDs and lasers that will be used for 3D sensing of humans and their environments.

Radiant ProMetric® Imaging Radiometers greatly simplify the process of measuring NIR light sources by capturing complete angular light source distributions in a single image for analysis. Users of Radiant NIR measurement systems benefit from turnkey solutions combining high-resolution imaging radiometers, optical components, and software test libraries for common analyses. Software tests in Radiant’s TrueTest™ Software platform include pass/fail parameters and custom output, and can be run in sequence to efficiently characterize and qualify light sources used for facial & gesture recognition, eye tracking, and other NIR-based 3D sensing applications.

radiant vision system wechat