Theory and simulation of calculating local illuminance density based on high dynamic range panoramic maps

Abstract

The development of the lighting profession toward the third stage requires our attention shifting from the light on certain planes to the light distributions in 3D spaces. In this article, we propose a practical strategy to measure the local density of illumination in 3D scenes based on the zeroth-order of spherical harmonics decompositions of a high dynamic range (HDR) panoramic map. The basic functional principle of deriving illuminance density from HDR panoramic maps was presented, and hereinafter named as illuminance panoramic. Illuminance panoramic was compared with Cuttle’s approximated illuminance scalar, which is essentially a physical approximation of the average illumination over a sphere. To verify the measurements, the average illuminance over a sphere, approximated illuminance scalar, illuminance panoramic, cylindrical illuminance, semi-cylindrical illuminance, and horizontal illuminance were simulated via a model (probe in a sphere). The results indicate that the measurement of density of illumination using HDR panoramic maps has well coincided with its definition (i.e., the average illuminance over a sphere) while other illumination values vary with how the probes are located. The measurement theories were later verified using six HDR panoramic maps of real scenes. This research provides confidence in developing applications in mobile phones by capturing HDR panoramic maps to measure the density of illumination in 3D spaces.

DeepLux: A data-driven method and benchmark for 3D illuminance maps estimation

“…A common approach to represent the lighting of a scene is to use high dynamic range (HDR) environment maps. Several works 8,13,14 have demonstrated different methods that can be used to estimate such HDR environment maps. Gardner et al 15 proposes one of the first methods for estimating HDR illumination from a single image by training a deep neural network (DNN) that predicts the location of lights in a scene from a single limited field-of-view photo on a large dataset of automatically annotated low dynamic range environment maps; the DNN is then fine-tuned on a small dataset of HDR environment maps to predict light intensities.…”

Effects of inter-reflections on the correlated colour temperature and colour rendition of the light field

“…The physical light field (the spectral power as a function of location, direction and wavelength) can be decomposed as the sum of qualitatively different components via spherical harmonics (SH). [33][34][35] Here we consider just the first two components since those are the main determinants of the modelling and (colour) contrast. [36][37][38][39] The zerothorder SH, the light density, is associated with the diffuse light-field component, namely the integration of the spectral power over the sphere.…”

Electronic design of a spatial low-light camera

“…Among them: EMCCD detector due to no anti-diffusion function, narrow operating temperature range, circuit volume, power consumption and complexity and other problems [8][9][10][11] , is not conducive to the integration and reliability of the camera.…”