基于改进的Zernike矩亚像素边缘提取算法外螺纹非接触测量  

Non-Contact Measurement for Outside Screw Based on a Modified Sub-Pixel Edge Detection Algorithm of Zernike Moments

作　　者：蔡艳[1] 叶连祥 孙大为[1] 田华 朱俊杰[1] 胡志云 

机构地区：[1]上海交通大学材料科学与工程学院,上海200240 [2]上海纳铁福传动轴有限公司,上海201315

出　　处：《上海交通大学学报》2014年第10期1468-1472,1478,共6页Journal of Shanghai Jiaotong University

摘　　要：为了进一步改善螺纹图像轮廓提取效果、提高螺纹参数的计算测量精度,在分析基于Zernike正交矩亚像素边缘检测算法的基础上,建立了边缘灰度阈值与一阶导数相结合的模型,提高了亚像素边缘点的定位精度,并在此基础上提出了计算螺纹中径和牙型角的新方法.结果表明,该方法对于克服工件表面水油污染、提高检测精度具有显著效果.In order to improve the measurement accuracy of and the optimized algorithm of characteristic parameters, a edge gray threshold was proposed based on the sub-pixel screw using the modified edge detection method first-order derivative model combined with the edge detection algorithm and the Zernike moments. A new method was designed to calculate characteristic parameters, including pitch diameter and thread angle. The results show that the modified algorithm has improved the precision and adaptive capaci ty of the measurement method.

关 键 词：螺纹检测 边缘检测 亚像素 Zernike正交矩 

分 类 号：TP274.5[自动化与计算机技术—检测技术与自动化装置]