机器人装配中的视觉引导定位技术研究  

Visual guiding and positioning technology of robotic assembly

作　　者：郑魁敬[1,2] 潘春莹[1,2] 许斐[1,2] 

机构地区：[1]燕山大学先进锻压成形技术与科学教育部重点实验室,河北秦皇岛066004 [2]燕山大学河北省并联机器人与机电系统实验室,河北秦皇岛066004

出　　处：《制造技术与机床》2016年第3期84-90,共7页Manufacturing Technology & Machine Tool

基　　金：国家自然科学基金资助项目(51275439)

摘　　要：机器视觉技术在工业自动化中的应用日趋广泛,在机器人装配过程中应用视觉技术可以提高装配的自动化程度。以花键轴套为研究对象,根据花键区域和非花键区域的明暗区别,采用区域生长的方法对花键区域进行提取。分析确定花键的几何特征点,利用双目测量模型进行三维测量,基于特征点进行图像匹配实现花键的精确定位。根据花键间隙进行花键对准检测,判别花键轴套的对准情况。由相机成像模型确定间隙所在像素区域,并作为检测窗口实现花键的对准。通过实验验证视觉引导定位技术在机器人装配中的可行性。The application of machine vision technology in industrial automation becomes more widespread. The vision technology in robotic assembly can improve the automation of assembly. The spline shaft and sleeve are studied. Depending on the brightness difference between the spline area and non- spline area,the regional growth method is used to extract the spline area. The geometric feature points of the spline are determined and the three- dimensional measurement is achieved based on the binocular model. Using the image matching method based on feature points,the precise positioning of spline can be achieved.The detecting alignment of spline shaft and sleeve is realized based on spline gap to discriminate the alignment. The pixel area of gap determined by the camera＇s imaging model is used as detection window and the alignment can be achieved. The feasibility of the visual guiding and positioning technology in robotic assembly is verified by the experiment.

关 键 词：机器视觉 区域生长 几何特征点 双目测量 图像匹配 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]