流线增强型线性积分卷积流场可视化  

Flow Visualization Based on Enhanced Streamline Line Integral Convolution

作　　者：韩敏[1] 张海超[1] 边茂松 郑丹晨[1] 

机构地区：[1]大连理工大学电子信息与电气工程学部,辽宁大连116023

出　　处：《系统仿真学报》2016年第12期2933-2938,共6页Journal of System Simulation

基　　金：国家重点基础研究计划(2013CB430403)

摘　　要：为全面清晰地展示出流场信息,减小线性积分卷积的计算量,提出一种流线增强型线性积分卷积方法。改进算法中流线生成部分:检测临界点位置并绘制渐变临界点区域,根据积分点所在区域结合不同的积分方法,并自适应更新积分步长。利用GPU的并行能力及使用GLSL(Open GL Shading Language),进一步增加输出图像的清晰度。仿真结果表明,该算法能够提高渲染速度同时保持可视化结果的清晰度,并利用改进的线性积分卷积方法对河流以及旋风流场进行可视化,以图像形式直观形象地展示流场分布信息。In order to show the flow information clearly and reduce the calculation of linear integral convolution, an enhanced streamline linear integral convolution algorithm was proposed. The generation of streamlines were improved. Critical points were detected and the points＇ area with a gradient fill was generated. Combined two integration methods in different areas, the integration step was updated adaptively. Utilizing the parallelism of GPU and GLSL （OpenGL shading language）, the algorithm further improved the sharpness of the output image. Experiments show that the improved linear integral convolution algorithm posses both the speedability and the precision. And the flow visualization of the river and hurricane with this improved method can show the flow field information intuitively.

关 键 词：流场可视化 线性积分卷积 临界点检测 GPU 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]