Large-scale Aerosols Visualization Based on the Accelerated Ray Casting Algorithm

Abstract

 The ray casting algorithm is a widely used basic volume rendering algorithm, but traditional ray casting algorithms have some disadvantages such as low efficiency, slow speed and large resource consumption when rendering aerosol data such as smoke, fog and dust in 3D scenes. In this regard, this paper proposes a large-scale aerosol visualization method based on a ray-casting acceleration algorithm. A large part of the computational work in a traditional ray casting algorithm is to compute a large volume of data from the underlying scalar field. Therefore, the accelerated design of the ray casting algorithm should first consider avoiding sampling of empty spaces and invalid voxels, and culling areas in the volume data that do not contribute to volume rendering. The method in this paper firstly optimizes and accelerates the traditional ray casting algorithm through the hierarchical bounding box tree, and combines the illumination model to enhance the color rendering of each voxel. Unnecessary space debris reduces hardware load, improves rendering speed, and meets the real-time rendering requirements of large-scale aerosols.