Optimization Of Massive 3D Model Rendering Based On WebGL

With the rapid development of computer graphics technology,three-dimensional graphics began to replace some of the application scenarios of two-dimensional graphics.Three-dimensional graphics have incomparable visuality of two-dimensional graphics,involving the medical field,the game field,the construction field,and the education field,and also bring the rendering pressure far beyond the two-dimensional image.Especially in recent years,the booming development of the video game industry has stimulated research on the rapid rendering of 3D scenes.Visibility culling is an important research direction for optimizing 3D scenes.There are many occluded and invisible objects in a large number of building models,and culling these objects can greatly reduce the amount of rendering.At the same time,because occlusion culling requires a large amount of computation,using the method of offline computing potential visible set to accelerate the calculation of occlusion culling in real-time rendering has become an effective means.In this regard,this paper proposes an algorithm based on scene management and ray casting to calculate the potential visible set.In this paper,the octree is used for scene management,and the 3D scene is divided into several scene units,and the potential visible set of each scene unit is calculated by projecting rays.In the experiment,the building model with the number of objects of about 50,000 is calculated.The potential visible set saved by each scene unit is no more than 14,000 objects,and the actual rejection exceeds 72%,which successfully reduces the amount of data rendered.This paper implements the three parts of database,background preprocessing and front-end real-time rendering.Accelerate real-time rendering with pre-computed potential visible sets.In the experiment,the rendering frame rate of the three-dimensional model with the number of objects of about 50,000 is increased from 4 frames per second to more than 14 frames per second,and there is no obvious object missing to affect the picture quality.This indicates that the rendering optimization of the large-scale 3D model can quickly complete the visibility culling under the premise of ensuring the picture quality,and improve the rendering frame rate of the 3D scene.