视觉检测中齿轮外轮廓分段方法研究

发布时间：2018-03-06 16:17

本文选题：视觉检测　切入点：齿轮分段　出处：《组合机床与自动化加工技术》2016年04期 　论文类型：期刊论文

【摘要】：在基于机器视觉的齿轮测量过程中,由于齿轮外轮廓各部分的形状、精度以及测量要求均不相同,需要对齿顶、齿廓和齿根分别进行数据处理,因此齿轮外轮廓分段至关重要。针对机器视觉系统测量视野小,单幅图像只能包含齿轮部分轮齿的问题,提出一种齿轮外轮廓分段方法。通过八邻域边缘追踪法提取齿轮外轮廓边缘,对该边缘进行最小二乘法圆拟合确定半径,并进行直线拟合确定齿轮中心逼近方向,经多次拟合后,逐渐逼近齿轮中心。齿轮中心确定后,实现对齿轮齿顶、齿廓和齿根部分的分段。实验结果表明,该方法可以确定齿轮中心,实现对齿轮外轮廓分段,程序简单,可操作性强,有利于提高系统的测量精度。
[Abstract]:In the process of gear measurement based on machine vision, because the shape, precision and measurement requirement of each part of the outer profile of gear are different, it is necessary to process the tooth top, tooth profile and tooth root separately. Therefore, it is very important to segment the outer profile of the gear. In view of the small measuring field of the machine vision system, a single image can only contain part of the gear teeth. In this paper, a method of gear outer contour segmentation is proposed. The outer contour edge is extracted by eight neighborhood edge tracing method, and the radius is determined by least square circle fitting, and the approaching direction of gear center is determined by straight line fitting. After many times fitting, the gear center is gradually approaching the gear center. After the determination of the gear center, the gear top, tooth profile and tooth root are segmented. The experimental results show that the method can determine the gear center and realize the segmentation of the outer profile of the gear. The program is simple and operable, which is helpful to improve the measuring accuracy of the system.
【作者单位】：沈阳工业大学机械工程学院;辽宁科技学院电气与信息工程学院;
【基金】：十二五国家科技支撑计划(2014BAF08B01)
【分类号】：TP391.41;TG86

【相似文献】