Analysis of the visibility of near-infrared illuminators

Abstract

What should be the maximum level of visible red light emitted from an infrared illuminator? This question has become relevant with the recent installation of active night vision systems on new vehicles. Such systems use IR-pass filters on powerful illuminators to reduce energy output in the visible range so that drivers of oncoming vehicles do not see them. Although eliminating output in the red range is possible, there is a practical desire to allow “leakage” of a limited amount of red light so that the effectiveness of the IR illumination is as high as possible, as long as it does not violate regulations. ECE regulations require that no red light can be visible at 25 m in front of the vehicle. In this analytic study, we developed a model based on CIE 19/2.1, equations from the PCDETECT model, and a median market-weighted headlamp intensity matrix to compute the threshold visible brightness of an infrared illuminator such that it would comply with ECE regulations. The model uses geometric parameters (position of the observer, illuminator, and glare sources), light source description (illuminator intensity and background luminance), and human performance data (visibility thresholds from PCDETECT, age, and population variance) to provide a spatial distribution of the threshold brightness values of an IR illuminator. The model predicts that in order not to detect a near-infrared illuminator, the illuminator should be positioned as close as possible to the low-beam headlamps. When positioned 0.02 m from the headlamp, the threshold for detection by a driver of an oncoming vehicle is predicted to be about 0.2 cd. We show that at 25 m in front of the car, if the IR illuminator has a U.S. high-beam pattern, it will first be detectable (in the visible red range) directly in front of the vehicle at around 1.2 m above the ground and 0.5 m left of the center.