Research On Direct Volume Rendering Related Technology

Volume rendering is one of the main methods in scientific visualization field. The internal structure information of volume datasets can be visualized interactively by volume rendering. Volume rendering is also known as direct volume rendering, which can visualize the original volume datasets without any intermediate geometry primitives, and generate the image of dataset with high quality details and context. Volume rendering has been applied in many fields, like medical, meteorology and geological domain.There are high noises in actual gathering datasets, which make the image of volume rendering not clear and smooth some structure features. The structure features can be damaged by simple filter methods suppressing the noise. To realize the feature enhanced visualization, we use gradient tensor to define the linear feature and adopt anisotropic diffusion filter algorithm to keep more structure features. We discuss our anisotropic diffusion filter algorithm with rand noise, pepper and salt noise with experiment.Direct volume rendering methods accumulated each voxel's optical properties along the casting ray, but which depend on the sample rates of dataset and transfer function. Traditional integration methods cannot get high quality image without enough sample rates. We implement pre-integrated volume rendering and peak finding volume rendering methods to get high quality visualization effect. We directly adjust the bounding box of dataset to realize the high quality clipping volume rendering. Lastly, we use local illumination model for volume rendering, which can strengthen the visual perception of visualization image.We integrated the above algorithm into an original system, which including rendering, feature enhancement and transfer function module and can be used in many different applications. After many datasets experiment, our volume system has good quality images and high efficiency.