摘要
针对传统的汽车冲压件端拾器的线上装配问题,提出了一种基于图像识别的线下安装方法。采用两个摄像机对装配中的端拾器进行视觉检测,通过图像预处理提取图中吸盘的轮廓,利用最小二乘法拟合其轮廓曲线,提取吸盘特征,并与模板库中的吸盘样本图像进行相似度匹配,根据吸盘的圆面特征和空间坐标变换原理,获取吸盘的位姿,通过计算其与理论位姿的偏差,给出调整的方法,为吸盘的正确装配提供指导。实验结果表明,该方法可以有效地提取吸盘的位姿,提高了安装吸盘的效率和精度。
To solve the problem of traditional on-line assembly of tooling of automobile stamping parts,this paper proposed an off-line installation method based on image recognition.Two cameras were used to detect the tooling in assembly process.The contour of suction cups was extracted by image preprocessing.The least square method was applied to contour curve fitting.Features of suction cups were then extracted and matched with the sample images in template library.The suction cup’s posture was obtained according to the characteristics of the circular surface and the spatial coordinate transformation.By calculating the deviation from its theoretical posture,an adjustment method was developed,which provides guidance for correct assembly of the suction cup.The results show that this method can effectively extract the suction cup’s posture and improve its installation efficiency and accuracy.
作者
陈坤煌
樊留群
杨永
CHEN Kunhuang;FAN Liuqun;YANG Yong(Sino-German School for Postgraduate Studies,Tongji University,Shanghai 201804,China)
出处
《机械》
2020年第5期60-66,共7页
Machinery
关键词
端拾器
真空吸盘
图像识别
位姿验证
tooling
vacuum suction cup
image recognition
posture verification
