摘要： 针对复杂流场中示踪粒子跟随性差、追踪困难等问题，提出一种新的虚拟粒子图像示踪方法，用于进行复杂流场分析。通过选定特定区域的图像作为虚拟示踪粒子，利用互相关算法追踪这些粒子的运动，以生动地表征粒子在图像内的运动情况。实验结果表明，采用该方法在流场进行示踪时，2帧图像间的误差大约在10-2像素量级，精度较高，适应性较好。

关键词: 数字图像, 互相关算法, 流场, 虚拟粒子, 示踪方法

Abstract: Searching tracking particles in complex fluid field always involves great difficulties due to its poor following performance. In order to solve the problem, this paper proposes a novel virtual particle image tracking method and applies it in the fluid field analysis. By selecting predetermined image blocks as virtual tracking particles, the movement of fluid can be vividly illustrated by tracking the movement of these virtual particles with image cross-correlation algorithm. Experimental results show that the tracking error in two successive images can be as low as 10-2 pixel, and the method has the merits of high accuracy and flexibility.

Key words: digital image, cross-correlation algorithm, fluid field, virtual particle, tracking method

中图分类号: 

• TN911.73