Multi-point simultaneous illuminance measurement with high dynamic range photography

Daylight illuminance uniformity is a common criterion when assessing the performance of fenestration and solar control. To evaluate a design according to this requirement, a grid of horizontal illuminance readings for a given time is required. In simulation based assessments, this can be calculated by defining sensor points at working plane level and running a simulation for a single time step, or an annual simulation to assess the uniformity for a typical year. Backing such predicted assessments with measured data has been a task involving installation and calibration of sensors, cabling and data acquisition systems, which typically render the assessed space unusable for the time of assessment. Using sequential readings from handheld devices as a convenient alternative ignores the dynamics of daylight, as the sky conditions cannot be assumed constant during the time required to record the illuminance at the required amount of locations.
We propose instantaneous image-based measurements of horizontal illuminance for assessments of daylight uniformity. Instead of cabled sensors, near-Lambertian reflectors are placed at working plane level in the assessed space. High Dynamic Range (HDR) images of the working plane with markers are taken and the luminance of the marker surfaces is extracted from the corresponding pixel values. These luminance values can be used to calculate horizontal illuminance for each marker location, assuming Lambertian reflection. As direct sunlight at working plane is considered to exceed the range of acceptable illuminance for uniformly lid spaces, only readings below a threshold of 3000 lm/m2 are considered.