微孔几何参数视觉测量研究  

The Research of Visual Measurement on Geometrical Parameters of Micro Circle Hole

作　　者：廖璇[1] 蔡勇[1] 廖晓波[1] 

机构地区：[1]西南科技大学制造科学与工程学院,四川绵阳621010

出　　处：《机械设计与制造》2014年第7期240-242,共3页Machinery Design & Manufacture

基　　金：高档数控机床与基础制造装备科技重大专项(2012ZX04007-021-07)

摘　　要：圆是视觉测量中重要的特征。针对圆形微孔几何参数的视觉测量进行了研究,提出了高效的圆参数测量与定位的测量方法。该方法根据分析采集图像的具体特性,通过对图像阈值分割、边缘提取等算法的研究,进而提取了目标的特征信息。面向检测区域小的尺寸测量问题,对比了改进的Hough变换法与最小二乘法,分析了引起测量误差的因素,通过最小二乘法实现圆形微孔的精确定位及直径等的精确测量。实验表明,系统的实际测量精度为0.005mm,该方法几何参数测量精确,定位准确,在视觉测量中具有广泛的应用前景。Circle is the important element in the computer vision measurement, The vision inspection method for the geometrical parameters of micro circle hole is studied. An efficient method of circle geometrical parameters and location is proposed. The available information of target object can be obtained by using the specific characteristic of the image with the search on image threshold segmentation and edge extraction. Compared with an improved Hough transform circle detection algorithm, it analyzed the cause of error in the measurement, the precision locating and diameter of micro circle hole can be recognized by using the least square fitting circle method, especially for the size measurement of a small region. Experimental results show that the measuring accuracy of the system is 0.005mm, and the method can obtain the geometrical parameters and the precision location, and can be extensively applied to the vision measurement.

关 键 词：视觉测量 微孔 特征提取 最小二乘法 

分 类 号：TH16[机械工程—机械制造及自动化]