文献类型：期刊文献

中文题名：基于机器视觉的齿廓偏差测量方法

英文题名：Measuring Method of Tooth Profile Deviation Based on Machine Vision

作者：曹鹏勇[1];王建文[1];程敏杰[1]

机构：[1]华东理工大学机械与动力工程学院,上海200237

年份：2021

卷号：21

期号：7

起止页码：2677

中文期刊名：科学技术与工程

外文期刊名：Science Technology and Engineering

收录：CSTPCD;;北大核心:【北大核心2020】；

基金：国家重点研发计划(2016YFF0203000)。

语种：中文

中文关键词：齿廓偏差;视觉测量;点云配准;边缘检测

外文关键词：tooth profile deviation;visual measurement;point cloud registration;edge detection

摘要：齿廓偏差是影响齿轮传动平稳性的重要参数。针对现有齿廓偏差测量方法的局限性,首先,以机器视觉为基础采集齿轮的图像,利用MATLAB进行图像处理获得清晰的齿廓边缘,采集齿廓目标点云数据。其次,运用逆向工程的方法建立标准齿轮模型获得齿轮的参考点云,结合经典迭代最近点算法实现齿廓目标点云和参考点云的配准,采集配准后的数据并进行曲线拟合,获得齿廓偏差并分析误差。最后,通过齿廓偏差计算值和测量值的对比和分析。结果表明,所提出关于测量齿轮齿廓偏差的方法具有合理性和可行性。
Tooth profile deviation is an important parameter affecting the stability of gear transmission.Aiming at the limitation of the existing tooth profile deviation measurement methods,the image of gear was collected based on machine vision firstly.The clear edge of tooth profile was obtained by image processing with MATLAB,and tooth profile target point cloud data was collected.Secondly,a standard gear model was established by reverse engineering to obtain the reference point cloud of the gear.The classic ICP algorithm was used to achieve the registration of the tooth profile target point cloud and the reference point cloud.After the registered data collected and performed curve fitting,the tooth profile deviations were obtained and analyzed.Finally,the calculated value and the measured value of the tooth profile deviation were compared and analyzed.The results show that the proposed method for measuring tooth profile deviation is reasonable and feasible.