Should DMS use 850nm or 940nm illumination?

The choice depends on sensor sensitivity, visible red glow, optical power, eye-safety design, packaging and algorithm needs. Lens and filter specifications must follow the selected system band.

What does NIR confocal mean for an RGB-IR lens?

It describes controlling the focus relationship between visible and near-infrared bands so both deliver useful image quality under the chosen focus strategy.

Can an IR-cut filter be selected independently from the lens?

It should be evaluated as part of the sensor-filter-lens stack because thickness, angle, spectrum and temperature can affect focus, color and transmission.

RGB-IR and 940nm Lens Design for DMS & OMS

1. Define the monitoring function and illumination mode

DMS may prioritize the driver face, eyes and head pose. OMS may need wider cabin coverage, occupant classification or child-presence regions. Combined systems create a different FOV and depth requirement.

State whether the camera operates in visible light, NIR only or RGB-IR fusion, and whether illumination uses 850nm, 940nm or another defined band.

DMS, OMS or combined coverage
Face box, eye box and cabin zones
Visible / mono NIR / RGB-IR mode
Illumination wavelength and geometry

2. Match spectral transmission to the complete stack

Lens material and coatings, filters, sensor quantum efficiency and illuminator spectrum determine the signal that reaches the image pipeline. A high lens transmission number at one wavelength does not describe the complete system.

Define useful passbands, blocking needs, angular response and temperature dependence. Review the filter at relevant ray angles because wide fields can shift spectral behavior.

Sensor spectral response
Lens and filter transmission
850nm / 940nm source spectrum
Angle and temperature dependence

3. Control visible-to-NIR focus and image quality

Optical materials disperse visible and NIR wavelengths differently. A lens focused for daylight color may lose NIR contrast, while an NIR-optimized lens may compromise visible color performance.

For RGB-IR, define MTF or image-quality targets in each relevant band and state the focus policy. Evaluate the interaction with sensor pixel architecture and demosaicing.

Visible-band image quality
NIR-band MTF / contrast
Visible-to-NIR focal shift
RGB-IR pixel and ISP behavior

4. Treat IR flare as a camera-system problem

NIR LEDs can create direct reflections, cover-glass ghosts, barrel scatter and sensor-stack reflections. Their position relative to the lens and cabin surfaces changes the artifact pattern.

Model and test representative illuminator positions, pulse conditions, eyewear and reflective cabin materials. Baffles, coatings, mechanical separation and ISP controls may all contribute.

LED position and emission angle
Cover-glass and sensor-stack reflections
Eyewear and cabin-surface reflections
Optical, mechanical and ISP mitigation

5. Validate coverage, eye box and production variation

A nominal camera must keep required face and cabin regions inside the usable image across occupant positions, seat adjustment, camera pose and manufacturing variation. Edge illumination and NIR contrast matter in the real geometry.

Validate multiple units, temperature conditions and representative cabin configurations. Link lens and module tolerances to algorithm confidence, not only to a single laboratory image.

Face / eye / cabin coverage
Camera pose and seat-position variation
Temperature and production samples
Algorithm-level confidence review

RGB-IR and 940nm lens design for DMS and OMS