Monte Carlo Sampling And Reusing Algorithm In Global Illumination Rendering

The global illumination rendering in 3D scenes produced by area light source is one of the important research topics in 3D graphics.The global illumination algorithms are mainly divided into three categories: ray tracing algorithm,radiosity algorithm and photon mapping algorithm.Among them,ray tracing algorithm is the more popular algorithm at present.However,in order to render the natural lighting effect from area light source well,it needs to generate lots of light source samples for per visible scene point,which can cause considerable computing overhead of visibility.In order to reduce the number of light source sampling points for Monte Carlo visibility sampling,a global illumination effect rendering algorithm based on adaptive iterative Monte Carlo visibility sampling is proposed.First,a binary area light source visibility map of 3D scene is created by casting shadow rays.Then,the penumbra heuristic information of the map is obtained using visibility gradient detection algorithm.Accordingly,the number of light source sampling points for per visible scene point is computed adaptively,and the area light source visibility of visual scene points can be calculated after two iterative Monte Carlo visibility sampling.The experimental results show that the proposed algorithm can render realistic global illumination effect with less light source sampling points.In order to obtain interactive speed,a global illumination effect rendering algorithm based on area light source visibility spatial reusing is proposed.First,an interleaving-based method of area light source sampling is designed.Then,an area light source-visible scene region visibility sampling map is generated by the method.Finally,the method of visibility filtering estimation using the characteristics of spatial correlation of visibility is suggested.Based on this method,the area light source visibility percentage of per visible scene point in the map is estimated.The experimental results demonstrate that the proposed algorithm can significantly reduce the calculation amount of area light source visibility and generate realistic global illumination effect at interactive frame rates.