A BRDF Statistical Model Applying to Space Target Materials Modeling

摘要

In order to solve the problem of poor effect in modeling the large density BRDF measured data with five-parameter semi-empirical model, a refined statistical model of BRDF which is suitable for multi-class space target material modeling were proposed. The refined model improved the Torrance-Sparrow model while having the modeling advantages of five-parameter model. Compared with the existing empirical model, the model contains six simple parameters, which can approximate the roughness distribution of the material surface, can approximate the intensity of the Fresnel reflectance phenomenon and the attenuation of the reflected light's brightness with the azimuth angle changes. The model is able to achieve parameter inversion quickly with no extra loss of accuracy. The genetic algorithm was used to invert the parameters of 11 different samples in the space target commonly used materials, and the fitting errors of all materials were below 6%, which were much lower than those of five-parameter model. The effect of the refined model is verified by comparing the fitting results of the three samples at different incident zenith angles in 0 azimuth angle. Finally, the three-dimensional modeling visualizations of these samples in the upper hemisphere space was given, in which the strength of the optical scattering of different materials could be clearly shown. It proved the good describing ability of the refined model at the material characterization as well.