Modeling And Simulation Of Infrared Reflectance Characteristics Based On The GPU Programming

The existing infrared scene simulation system mainly focuses on the intrinsicradiation. It is lack of fine modeling and simulation for environmental radiationreflected from the target, which has seriously affected the authenticity of the simulationscenarios, and also restricts its application in the military field such as the design andevaluation of the infrared imaging guidance weapons and the infrared targetidentification and tracking guidance training.In this paper, the modeling and simulation of the reflection characteristics in thereal-time infrared scene are explored and studied. The simulation of the environmentradiation’s reflection from target in the medium-wave and long-wave is realized basedon the GPU programming. Firstly, starting from the principle of imaging detectors,infrared reflection theory is introduced. By Modtran calculation, analysis of theenvironment radiation characteristics in the medium-wave band and long-wave band aremade. Secondly, by comparative study of a variety of reflection models, Schlick BRDFmodel is recommended to be feasible in real-time simulation. Based on analyzing theaffection of reflective properties and environmental radiation in the simulation throughexperiment and theoretical calculation, a band reflectance calculation model isestablished. Then a method of reflective properties simulation is made based on GPUprogramming, and the reflection model used in this paper is verified by calculation andsimulation. Finally, on the OGRE-based infrared scene simulation platform, thesimulation of the infrared reflection characteristics is achieved, which improves thissimulation framework. To meet real-time simulation, the environment radiation,material parameters, integral data and the other pre-generated data are preserved in theform of texture. The GPU fast sampling greatly solves the problem of frame ratedropping after including the calculation of reflection. The simulation results show thatthe method effectively increases the simulation authenticity under the premise ofslightly decreasing the simulation efficiency.