线结构光圆柱度测量装置的转轴标定算法

Rotary Axis Calibration Algorithm of Cylindricity Error Measuring Device Using Linear Structured Light

利用机器视觉来实现零件尺寸的测量因其操作方便,节省人力,越来越受到学者的重视。利用结构光来实现轴类零件的测量时,需要预先知道回转轴线在相机坐标系的位置。首先回顾分析了目前提出的转轴标定算法,讨论了以往的标定方案的可取和不可取之处,然后在前人基础上提出了一种新的解决方法。首先通过夹具将标定板固定在三爪卡盘上,使标定板相对三爪卡盘空间位置固定,然后旋转三爪卡盘从而得到一系列标定板图像。因为标定板和轴心有确定的几何关系,通过这一系列标定板平面上的特征点通过最小二乘法即可演算出轴心位置。在所提方法的基础上进行了实验,结果表明提出方法准确且具有一定应用价值。

Taking advantage of machine vision to measure the size of component has attracted more and more attentions due to its convenience in operation and labor saving.When using structured light to measure a shaft part,it is necessary to know the position of the rotary axis with respect to the camera coordinate system.This paper first reviews the methods proposed before and discusses the advantages and disadvantages of those methods.Then,on the basis of the predecessors,this paper proposes a new solution.Firstly,the calibration plate is fixed on the three-jaw chuck to make the plate fixed relative to the axis,and then the calibration plate is rotated to obtain a series of images.Because the calibration plate and the axis are determined by the geometric relationship,the axis position can be calculated by the least square method through the feature points on the plane of the calibration plate.Experiments show that the proposed method is feasible and accurate.