Research And Implement On Volume Rendering For Seismic Data Based On GPU Ray Casting Algorithm

With the development of 3D visualization technology and the increase of the scale of 3D seismic data,in seismic field,the application of visualization technology has new challenges because of the scale of 3D seismic data and the complexity of visualization algorithm.On the volume rendering of 3D seismic data,since the volume data can not be completely loaded into the memory,preprocessing and data scheduling and rendering will be hindered.Therefore,the related algorithms have been studied and improved.After researchs,the construction algorithm of octree tree structure based on Morton code and the rendering algorithm of multi-resolution seismic data based on GPU are designed,so the process of 3D seismic data volume rendering is optimized.The construction of the octree tree based on the Morton code is the preprocessing process of the data before the volume rendering.After the algorithm is improved,the original seismic data is partitioned.Then the Morton code is used to code the blocks,so that every eight consecutive blocks can generate their father block.Finally the octree format seismic data is generated after the multi-resolution model octree processes blocks.This new seismic data is the source data for the seismic data.The rendering algorithm of multi-resolution seismic data based on GPU is an improvement of the traditional ray casting algorithm.The improved algorithm mainly improves something about the multi-resolution data.In the case of preserving the color synthesis in the traditional algorithm,the resampling method and the volume texture generating is improved,which makes the algorithm realize rendering multi-resolution volume data.In addition,the algorithm is transplanted into the GPU to calculate,so it greatly improves the efficiency of the implementation of the algorithm.At last,two sets of experiments are carried out to verify the optimized results of the two improved parts of 3D seismic data volume rendering.The results show that the preprocessing algorithm can be used to any kind size of the seismic data,and it can give some reference parameters.The multi-resolution ray casting algorithm based on GPU correctly displays the seismic data,and the image frame rate can be controlled within the allowable range with a reasonable data scheduling.