基于OpenCV的六角钢坯尺寸测量研究  

Research on Dimensional Measurement of Hexagonal Billet By OpenCV

作　　者：黄凯[1] 余学才[1] 唐飞[1] 丁建平[1] 李亚南[1] 

机构地区：[1]电子科技大学光电信息学院,成都610054

出　　处：《计算机测量与控制》2013年第5期1137-1139,共3页Computer Measurement &Control

摘　　要：在现代化工业生产中,非接触式钢坯尺寸测量已经成为钢坯检测中的一种十分重要的测量方法;设计了一种基于计算机视觉的六角钢坯尺寸自动测量系统,通过工业数字摄像头采集由激光器照射的钢坯图像,在OpenCV开源平台下,对其进行相关的图像预处理,利用霍夫变换识别出直线段,并提出了一种简单,有效的直线合并算法,最后实现对六角钢坯尺寸的智能测量;实验结果表明,此算法能准确快速地获得钢坯的尺寸信息,在角度测量时最大误差为0.9°,具有很好的精确性和实时性。The method of non--contact measurement to measure the size of billets has become a very important method to the detection of billets in the modern industrial production. A system of auto--measurement of hexagonal billets size was designed based on computer vision. In the open source workbench of OpenCV, the billets images were preprocessed and collected by industrial digital camera with laser device illuminated. Using Hough transform to identify a line segment, a simple and valid algorithm of line combination was raised to realize smart measurement of hexagonal billets size finally. The experimental results show： the accuracy and real--time size information of the measured billet can be obtained accurately and rapidly by using the algorithm. The maximum error range in the measurement of the angle is 0. 9°.

关 键 词：钢坯检测 OpenCV预处理 霍夫变换 直线合并 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]