空间矩定位齿轮齿距视觉检测  

Machine Vision for Axial Pitch Detection Based on Spatial Moment Location

作　　者：师平[1] 张丽鹏[2] 

机构地区：[1]江苏航空职业技术学院,江苏镇江212134 [2]江苏大学机械工程学院,江苏镇江212013

出　　处：《机械设计与制造》2016年第12期205-208,共4页Machinery Design & Manufacture

摘　　要：针对齿轮的精度测量问题,采用基于机器视觉的非接触检测方法,运用MATLAB软件中图像处理模块,实现了对圆柱齿轮齿距的测量。通过对图像的预处理,在运用Canny算子提取齿轮边缘的基础上,用空间矩边缘检测法,计算出一维阶跃边缘的前三阶空间矩,得出实际模型中边缘位置,提取出齿轮亚像素边缘。对比了两种齿轮中心定位方法,完成了对齿轮中心的精确定位。通过构造虚拟圆,获取分度圆上齿轮齿距。所提出的基于机器视觉的齿轮齿距检测方法具有很好的应用前景。Aiming at the problem of the secure detection for the gear,the detecting methods of non-contact based on machine vision were adopted.With the help of MATLAB image processing model,the detection of axial pitch was put forward.After doing the image preprocess,the gear edge was acquired by using Canny algorithm and the spatial moment edge detection method.the first three spatial moment of the one dimensional step edge was calculated to obtain the edge position of the actual model of gray value image,and the sub-pixel edge of the gear was also acquired.To complete the detection of axial pitch,two locating ways of the center of the gear were compared.Finally the accurate location of the gear center was worked out.By constructing a virtual circle, The axial pitches on the reference circle were calculated.The axial pitch test method based on machine vision proposed in this paper has a fine application prospect.

关 键 词：MATLAB 图像处理 齿距检测 

分 类 号：TH16[机械工程—机械制造及自动化] TP242[自动化与计算机技术—检测技术与自动化装置]