摘要
为达到非接触式测量的目的,采用机器视觉的方法来实现对运动机构的运动参数的测定。通过在机构上设定一个标志,利用模板匹配技术,对该目标实现实时在线跟踪,得到它在动态图像中每次的坐标值,通过对动态图像的分析和处理从而在线得到各种运动参数的曲线。模板匹配算法采用了全局匹配与小区域匹配相结合的策略。小区域匹配一次所用时间为20 ms,完全能达到在线实时跟踪测试的要求。
The purpose of this paper is using the method of machine vision to measure the motion parameter of motion framework, consequently get the purpose of non - touching measure. By setting a sing on the framework and using pattern template matching technique to achieve the online real- time track of the sign and get the coordinate of the dynamic image of the sign. By analyzing and handling the dynamic image, get curves of several motion parameters. Pattern matching arithmetic adopts a method of combining global match and small - region match. Each time, the small - region match spend 20 ms,which can achieve the require of online real - time tracking test.
出处
《机械与电子》
2007年第10期49-52,共4页
Machinery & Electronics
关键词
机器视觉
图像识别
运动参数
模板匹配
machine vision
image matchingmotion parameter
pattern matching
