A More Robust Method for Digital Video Camera Calibration for Luminance Estimation 2022-01-0802

Mapping the luminance values of a visual scene is of broad interest to accident reconstructionists, human factors professionals, and lighting experts. Such mappings are useful for a variety of purposes, including determining the effectiveness and appropriateness of lighting installations, and performing visibility analyses for accident case studies. Previous work has shown that pixel intensity captured by consumer-grade digital still cameras can be calibrated to estimate luminance [1-7]. Taking a digital still image and converting this image into a luminance map even further reduces the time required for luminance measurement. Suway and Suway previously presented a methodology for estimating luminance from digital images and video of a scene [1]. In this paper, the authors update this methodology for calculating luminance from a digital camera. The updated calibration method results in more accurate luminance estimation over the entire range of the camera’s sensor, particularly in low-light conditions. Further refinements to the fitting methodology result in a robust calibration even in the presence of measurement noise. Ultimately, the presented methodology allows for the user to mount a camera near the driver’s eye location and to drive through a scene capturing video or still images. Still images from the video are then exported and analyzed, creating a luminance map. The method previously presented by Suway and Suway [1] is compared to the updated method. It is shown that throughout the range of pixel values, the updated method accurately estimates the luminance values. In many cases, the results are comparable to the previously published method [1]; however, the updated method is more robust at the extremes and more robust to noise.