基于视觉引导的自动组装技术探究  

Research on Automatic Assembly Technology Based on Visual Guidance

作　　者：陈发毅 贺敏 吴蕾 秦付军[1] 何紫杨 CHEN Fayi;HE Min;WU Lei;QIN Fujun;HE Ziyang(College of Mechanical Engineering,Xihua University,Chengdu 610039,China;Suzhou Funa Intelligent Technology Co.,Ltd.,Suzhou 215000,China)

机构地区：[1]西华大学机械工程学院,成都610039 [2]苏州富纳智能科技有限公司,江苏苏州215000

出　　处：《机械工程师》2022年第1期60-62,共3页Mechanical Engineer

摘　　要：机器视觉技术在自动化领域应用广泛,传统的视觉组装大多基于单个相机拍照、抓取、校正、组装的步骤来进行,忽略了抓取后的位置信息,使得组装的精度不高。设计四轴直角坐标型机器人控制系统,引入下相机对产品进行二次定位。对上下相机进行手眼标定,通过视觉处理引导机械手完成组装,为验证系统的有效性,以方块组装进行试验,在取料、放料位置和角度未知的情况下,对方块进行位置和角度补偿。视觉定位引导系统的定位精度能达到0.192 mm,实验结果表明基于下相机的视觉组装具有可行性。Machine vision technology is widely used in the field of automation. Traditional vision assembly is mostly based on the steps of a single camera taking pictures, grasping, correcting and assembly, ignoring the position information after grasping, resulting in low assembly accuracy. This paper designs a four-axis rectangular coordinate system robot control system, and introduces the lower camera to perform secondary positioning of the product. The upper and lower cameras are calibrated by hand and eye, and the robot is guided to complete the assembly through visual processing. In order to verify the effectiveness of the system, the test is carried out with block assembly. When the position and angle of the reclaiming and unloading are unknown, the position and angle of the block are compensated. The positioning accuracy of the visual positioning guidance system can reach 0.192 mm, and the experimental results show that the visual assembly based on the lower camera is feasible.

关 键 词：机器视觉 直角坐标机器人 定位引导 精确组装 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]