Research On The Statistical Sampling Method Of Henyey-Greenstein Phase Function

Monte Carlo radiative transfer(MCRT)algorithm is often used to simulate the multiple scattering process of light in natural media.It has been widely used in computer graphics rendering and astrophysics.An important step in Monte Carlo radiative transfer is to sample the scattering angle of the phase function.In this paper,Monte Carlo method is used to calculate the scattering angle in order to explore a suitable and efficient sampling method to simulate light scattering.First of all,we discuss the research on the statistical sampling method of Rayleigh phase function.We use three sampling methods to sample the Rayleigh phase function,including the rejection sampling,direct sampling and tabulation method,and apply the three sampling methods to Henyey-Greenstein(HG)phase function.In order to simulate the scattering characteristics in different media,we improve the HG phase function to obtain the Rayleigh Henyey-Greenstein(RHG)phase function and Modified Henyey-Greenstein(MHG)phase function.The sampling method is used to sample the improved phase function,and the spatial distribution of each scattering phase function can be seen according to the rejection sampling and direct sampling.Moreover,in order to reduce the sampling error and cost,we consider the combination of tabulation method and rejection sampling to propose the improved tabulation method.In addition,we use Kolmogorov-Smirnov(KS)test to test the sample points to analyze the effect of sampling method.According to the results of KS test,we discuss and explain the sampling effect,advantages and disadvantages of three sampling methods and improved tabulation method.Finally,different g values are selected to prove that the parameter g has a significant impact on the accuracy of the sampling method.