Research On Parallel Algorithm For Vector Field Visualization Based On Texture

Visualization in Scientific Computing (ViSC) is a new subject that was brought forward and developed rapidly in the late 80's. It utilizes advanced technology and method of computer graphics and image, to excavate the information implied in the data field of space, and to change it into the computer image that human vision can be perceived. It makes researchers understand information more rapidly and accurately. Vector field visualization is one of the most challenging issues in ViSC. It changes the extensive vector data emerging in science and engineering computing into figures and images, and displays them. It makes people explain and understand objective law included in the abstract science data visually. Thus it has extensive application.Line integral convolution (LIC) is one of texture-based methods for vector field visualization. Even in the vector direction changing sharply fields, the method can show the direction of the vector field well, and reveal details of the vector field better. In addition, it can fully disclose the relativity among every spot of the vector field. The image generated by this method is continuous and fine. LIC has broad development prospects, but it demands much running time, what limit its utilization.In recent years with the development of parallel computing, it has already become the most effective and most popular method to improve computing capability. In narrow sense, parallel computing is running a work, which is running formerly by sequence, simultaneously in the parallel mode on several processors. The aim of parallel computing is enhancing the efficiency of running a program largely.This thesis has carried on analysis and discussion to vector field visualization and parallel computing respectively at first. Then, we have introduced the most popular parallel computing environment MPI (Message Passing Interface). After that, we have carried on deeper research on LIC, and have combined thought of parallel computing and computation of LIC, and present parallel algorithm for LIC consequently. At last, under the distributed programming environment MPI, we have implemented the LIC parallel algorithm on the Dawn TC4000L computer cluster. The tests show that this parallel algorithm has a good performance and has certain realistic significance.