基于最小二乘法的螺纹图像检测研究  

Research on thread image detection based on least square

作　　者：孙珂琪[1] 黑棣[1] Sun Keqi;Hei Di(Shannxi Railway Institute,Shaanxi Weinan,714000,China)

机构地区：[1]陕西铁路工程职业技术学院,陕西渭南714000

出　　处：《机械设计与制造工程》2022年第7期86-90,共5页Machine Design and Manufacturing Engineering

基　　金：陕西省自然科学基金资助项目(2019JQ-928)。

摘　　要：利用图像检测技术检测螺纹参数时,从图像中提取清晰准确的螺纹轮廓是检测的基础。首先从图像获取误差和随机检测误差两方面分析了图像检测中误差产生的原因,基于LabVIEW图像处理模块利用图像增强数据、图像过滤、边缘检测算法对图像进行预处理;然后通过计算确定标定系数、设置坐标系,实现像素坐标系到真实坐标系的转换,在MATLAB中利用最小二乘法对螺纹图像的牙型进行重构、中径进行拟合;最后以M14圆柱外螺纹为例,将所提方法的螺纹参数检测结果与实测值和标准值进行比较,结果显示,螺距和中径误差在0.0015μm以内,牙型角误差在0.01°以内。When using image detection technology to detect thread parameters,extracting a clear and accurate thread profile from the image is the basis of detection.Firstly,the causes of errors in image detection are analyzed from the two aspects of image acquisition error and random detection error.Based on the LabVIEW image processing module,the image is preprocessed with image enhancement data,image filtering,and edge detection algorithms;Set the coordinate system to realize the conversion from the pixel coordinate system to the real coordinate system.In MATLAB,the tooth shape of the thread image is reconstructed and the pitch diameter is fitted by the least square method.Finally,taking the M14 cylindrical external thread as the research object,the test results of this method are compared with the measured values.The results show that the error of pitch and pitch diameter is kept within 0.0015μm,and the error of tooth profile angle is kept within 0.01°.

关 键 词：图像检测 最小二乘法 中径拟合 牙型拟合 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]