Real-time Translucent Material Rendering With Multiple Light Sources Based On Virtual Light Source

Subsurface scattering is an Indispensable part of realistic rendering,which enables the rendering of translucent objects with a sense of softness and fullness,and it can also show a sense of transparency in thinner places.The essence of subsurface scattering is the phenomenon of light entering the object from the point of incidence and then scattering away.Narrowly speaking,it is called transmission phenomenon when the exit point of the scattered light and the incident point is not on the same side of the object,and vice versa is subsurface scattering in the narrow sense.In the field of real-time rendering,the transmission of translucent objects is very closely related to subsurface scattering in the narrow sense: the transmission attenuation function for transmission rendering is evolved from the diffuse attenuation function of subsurface scattering in the narrow sense,and the outgoing transmitted light also needs to be diffusely attenuated by applying a subsurface scattering algorithm.The transmission thickness needs to be solved first which is the major difference between transmission rendering and subsurface scattering rendering in the narrow sense.This paper takes the calculation of transmission thickness in the rendering of translucent objects as the research object,combining and improving the fragment sorting technology,effectively reducing the transmission thickness calculation errors caused by light tangency and model irregularities.When it is used in multi-light source scenes,this method significantly reduces the consumption of video memory,and finally,when the scene is more complicated,it can also improve efficiency.The core idea of the algorithm in this paper is to set up a virtual light source in the scene,sort and store the depth information of the scene in the light source space.When calculating the transmission thickness at other real light sources in the scene,the step of sorting and saving fragments is no longer performed,but the connection between the real light source and the point to be shaded is sampled,and the ratio of the sampling point inside the object is calculated and multiply the length of the entire line segment to estimate the transmission thickness.Sampling will not bring any video memory overhead,but it will bring a certain efficiency overhead.Experiments have shown that when the scene is more complex,the efficiency overhead of sampling will be smaller than the sorting overhead.At the same time,multiple sampling points can also increase the error tolerance rate of thickness calculation,even if the light is tangent to the object,better results can be obtained.The main work of this paper is as follows:· Designed and realized the selection of the best virtual light source: the transmission thickness calculation of all real light sources in the scene will be multiplexed to the depth information at the virtual light source.It must be ensured that the complete and most accurate depth information of the scene can be obtained from the virtual light source.Besides,it is necessary to ensure that the virtual light source is minimally affected by the tangency of the light and the irregularity of the model.Therefore,this article uses the scene bounding box and considers the above factors to obtain the best virtual light source selection scheme.· Designed and implemented the real light source transmission thickness estimation algorithm: to avoid the video memory overhead at the real light source and the efficiency overhead of sorting fragments,this article no longer sorts and saves the scene fragment depth information at the real light source,but uses sampling For each ray and determine the proportion of the ray inside the object to estimate the thickness.The principle of judging whether the sampling point is inside the object comes from the ray method in computational geometry,and the depth information generated at the virtual light source is used to assist in the determination.· Optimizing sampling to further improve efficiency: In complex scenes with multiple light sources,the algorithm in this paper can achieve certain advantages in efficiency,but in scenes with fewer light sources,it will bring efficiency disadvantages due to multiple sampling.This article adopts adaptive adjustment of sampling Three strategies such as step size,reducing the resolution of depth information,and adding additional scene information at the real light source for auxiliary sampling have improved the efficiency of the algorithm.Based on the above content,and combined with the experimental results,it is shown that the algorithm in this paper combines the fragment sorting technology and improves it.Compared with the commonly used depth map algorithms in the industry,huge progress can be made in effect.Compared with the algorithm that only combines fragment sorting technology,the algorithm in this paper has certain advantages in effect for simple scenes with a single light source;for complex scenes with multiple light sources,the algorithm in this paper has certain advantages in three aspects: effect,efficiency,and memory consumption.