Footprinting-based Three-dimensional Seismic Data Parallel Visualization Methods

Visualization of Scientific Computing is the primary method in interpreting 3D seismic data. Using Computer Graphics and Image processing techniques, 3D seismic data can be converted to graphics or images displayed on screen for interactive processing. The main methods include surface visualization and volume rendering. With the volume rendering technology, overall images of 3D seismic data can be generated as well as every detail. The method has advantages as high quality images and easy for parallelization. However, it has to face with high calculated amount and long processing time.With the development of numerical simulation and measurement technology in all applications, the amount of 3D data grows rapidly. Therefore, how to improve the speed of volume rendering has become an urgent problem in Visualization research, parallel visualization technology is a feasible method.After described the principles of the basic footprint method and realized it firstly, we analyzed the problems of this algorithm and then put forward a improved method based on View-Buffer. The results of experiment shows that our new method can resolve the image blurring problems with original method effectively and can generate clearer image with higher-quality. After analyzed parallel computing process of our program, a parallel algorithm based on footprint is designed and realized on a computer cluster system for testing 3D seismic data. Compared with the serial algorithm, the rendering speed of parallel algorithm increased greatly.