“Efficient Visualization of Light Pollution for the Night Sky” by Dobashi, Ishikawa and Iwasaki
Conference:
Type(s):
Title:
- Efficient Visualization of Light Pollution for the Night Sky
Session/Category Title: Rendering, Neural Fields & Neural Caches
Presenter(s)/Author(s):
Abstract:
The artificial light sources make our daily life convenient, but they cause a serious problem called light pollution. We propose a system for efficient visualization of the light pollution for the night sky. A number of methods have been proposed for rendering the sky, but most of the methods focus on the rendering of the daytime or the sunset sky where the sun is the only light source. For the visualization of the light pollution, however, we have to take into account many city light sources on the ground, resulting in expensive computational cost. We address this problem by precomputing a set of intensity distributions of the sky illuminated by a city light for different locations and atmospheric conditions. We apply the principal component analysis and the fast Fourier transform to the precomputed distributions, allowing us to efficiently visualize the light pollution. We achieve one to two orders of magnitudes faster computation compared to a naive approach that accumulates the scattered intensity for each viewing ray. This fast computation further allows us to interactively solve an inverse problem that determines the city light intensity to reduce the light pollution. Our system provides the user with the forward and the inverse investigation tool of the light pollution.
