Combining adaptive illumination systems with computer vision allows for optimized lighting in various scenarios. In automotive applications, matrix configurations of individually dimmable LED pixels enable automated and adaptive lighting, detecting oncoming traffic and activating specific areas accordingly. Such systems also have potential for general lighting, where cameras detect objects and define a target illuminance distribution. However, illumination applications require smooth light output while realizing the target distribution. This paper proposes a least squares-based method for LED array pixel addressing, resulting in smooth and effective illumination.
|