摘要
针对活塞杆制造过程中同轴度检测出现的精度低,效率低问题,提出了基于机器视觉技术的活塞杆同轴度误差检测方法,并建立了同轴度误差数学模型。采集活塞杆图像并进行边缘特征提取;将基准圆柱表面分解成240个连续的横截面,待测圆柱表面分解成110个连续的横截面;基于最远Voronoi图计算出每个截面的最小外接圆的圆心,并用最小二乘法拟合出活塞杆圆柱面的轴线。以此方法获得基准轴线,求得零件的同轴度误差,并以某型号的活塞杆进行实验分析,结果与三坐标测量机(CMM)测得的误差结果相吻合,表明该方法可以有效、正确地进行同轴度误差的评定。
For low precision and inefficiency in detecting the coaxial error of piston rod,a method to inspect the coaxial error of piston rod based on the machine vision is presented. This method is proposed in terms of the principle of coaxial error,and then a model is developed. The piston image is acquired and the edge feature is extracted; the reference cylindrical surface is decomposed into 240 continuous cross-section,the testing cylindrical surface is divided into 110 continuous cross-section; the circum circle center of each cross-section based on the furthest Voronoi diagram is calculated,and the method of least squares fitting the axis of rod cylindrical surface is used. Datum axis are obtained in this way,thus the coaxial error of piston rod can be calculated. Experimental results of a certain piston rod showed that the present method is effectively evaluated and it is very close to the result measured by Coordinate Measuring Machining( CMM).
出处
《机械科学与技术》
CSCD
北大核心
2015年第12期1846-1850,共5页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(51075362)
浙江省自然科学基金项目(Y1100073)
宁波市产业技术创新及成果产业化重点项目(2013B10022)资助
关键词
机器视觉
活塞杆
同轴度误差
VORONOI图
calculations
calibration
CCD camenas
computer vision
efficiency
errors
feature extraction
flowcharting
image acquisition
inspection
least squares approximations
mathematical models
matrix algebra
pixels
coaxial error
machine vision
piston rod
voronoi diagram
