Modeling atmospheric effects in the infrared scene generator with polynomial fitting

Abstract

OpenGL is widely used for the real-time infrared (IR) scene simulations. Atmospheric effects are one of the most important factors that affect the physical realism of the IR scene. Modeling atmospheric effects in the IR band using OpenGL's fog model approach is found in the literature. However, the related fog model needs single equivalent parameters for the atmospheric transmittance and atmospheric path radiance that are both range and wavelength dependent parameters. In this paper, a new model is proposed for the atmospheric transmittance and atmospheric path radiance using the least squares estimation with polynomial fitting by defining additional capabilities on the standard pipeline of OpenGL. Modeling of the atmospheric effects in the IR scene generator is performed by using the related model parameters in the shader program, which is built with OpenGL Shading Language (GLSL). Atmosphere data obtained with Modtran software are used to evaluate the performance of the proposed method. It is shown that the proposed model has an error less than 2% and decreases the GPU memory usage significantly.