摘要
为提高齿轮测量精度,提出一种基于组合形态学边缘检测算法的齿轮参数测量方法。对工业相机采集的齿轮图像做畸变矫正后进行滤波和二值化等预处理;针对传统边缘检测算法存在的不足,采用检测精度高、抗噪声性强的组合形态学边缘检测算子提取齿轮单像素边缘;采用连通域操作填充齿轮轴孔,利用质心法定位齿轮圆心;计算齿轮边缘点到圆心的距离;利用量块进行像素当量标定,转换后得到齿轮几何参数测量值。实验表明,对比几种边缘检测算法的测量精度和速度,所提出的齿轮参数视觉测量方法具有更高的测量精度和良好的抗噪性。
In order to improve the precision of gear measurement,a gear parameter measurement method based on combined morphological edge detection algorithm is proposed.The gear image collected by industrial camera was corrected for distortion and then preprocessed by filtering and binarization.For the shortcomings of the traditional edge detection algorithm,a combined morphological edge detection operator with high detection accuracy and strong noise resistance was used to extract the single pixel edge of gear.The gear shaft hole was filled by connected domain operation,and the center of the gear was located by centroid method.The distance from the edge of the gear to the center of the gear was calculated.The pixel equivalent was calibrated by measuring block,and the measured value of gear geometric parameters was obtained after conversion.The measurement accuracy and speed of several edge detection algorithms were compared.The experiments show that the proposed visual measurement method of gear parameters has higher measurement accuracy and good noise resistance.
作者
朱浪
方成刚
粟序明
Zhu Lang;Fang Chenggang;Su Xuming(School of Mechanical and Power Engineering,Nanjing Tech University,Nanjing 210000,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2023年第2期33-39,共7页
Computer Applications and Software
基金
国家自然科学基金项目(51635003)。
关键词
齿轮测量
机器视觉
形态学边缘检测
畸变矫正
Gear measurement
Machine vision
Morphological edge detection
Distortion correction