Lighting Systems

Mapping Illuminance Values Using HDR Imaging on Turf

Graduate Student: Alayne Nyboer (MSArE ‘21)
Advisor: Arpan Guha, Ph.D.

For the graduate technical project, graduate student Alayne Nyboer explored the viability of creating illuminance maps from High Dynamic Range (HDR) images for turf-based exterior surfaces.

In the recent past, the photorealistic nature of HDR images have been utilized to extract photometric quantities from the image, with luminance being the most mapped quantity for both interior and exterior settings. The accuracy of mapping luminance values of a scene from HDR images has been validated in recent research studies. On the contrary, extraction of illuminance values in a scene has been less common, presumably because of the complex nature of the calculation as well as the potential presence of confounding variables such as objects with significant specular components. Mapping illuminance for exterior HDR scenes has been studied minimally. For indoor scenes, the feasibility of illuminance mapping from HDR images has been explored but most studies have several conditional requirements regarding surface properties and overall light-levels in the mapped scene, most importantly, the proposed methods seem to work better on predominantly diffuse surfaces or surfaces without significant specular components. Grass and natural vegetation have been shown to exhibit near-perfect diffuse behavior at viewing angles greater than 75° and past research has suggested that it is very reasonable to assume most types of soil or dirt exhibit diffuse characteristics when dry.

Data was collected at the Blue Devil’s Stadium, Southfield MI under floodlights. Analyses showed an average of 5.6% error across the illuminance map created from the HDR images (versus manually measured illuminance values). This may have major implications in lighting measurement as the traditionally time-consuming and labor-intensive illuminance measurement process for sports stadiums can now be potentially replaced by mapping illuminance from HDR images. A drone flown above the facility can capture images that can later be run through post processing to retrieve illuminance measurements of the entire field.