摘要: |
针对航空航天等领域蜂窝夹层结构件机器人化制造中的蜂窝型特征精确定位问题,提出了一种利用双目视觉引导工业机器人进行蜂窝夹层结构件单孔灌注的定位方法。目标区域分割原始双目图像,并通过双目匹配计算生成三维深度图;边缘检测、膨胀、腐蚀、骨骼提取、多边形拟合处理左图像蜂窝棱边,获得蜂窝角点特征的二维坐标,再利用深度图获得蜂窝特征角点空间坐标。利用6个角点坐标计算得到蜂窝芯孔的空间定位位姿信息。最后,在工业机器人和双目相机构成的定位系统上定位试验蜂窝结构件。结果表明:该定位方法能够精确定位存在一定变形的蜂窝芯孔(非标准六边形),其定位精度约为2mm,能够满足灌注作业精度要求。 |
关键词: 工业机器人;机器视觉;机器人定位;蜂窝特征 |
基金项目:国家自然科学基金(51805071);中央高校基本科研业务费DUT17JC16、DUT18RC(3)073。 |
|
Robot Vision Location Method for Honeycomb Structures |
Zhang Jiali1, Li Te1, Tan Chaoyuan2, Liu Haibo1, Liu Kuo1, Li Lanzhu2, Wang Yongqing1
|
1.Key Laboratory for Precision and Non-Traditional Machining Technology of Ministry of Education;2.Aerospace Research Institute of Materials & Processing Technology
|
Abstract: |
Aiming at the problem of accurate localization of honeycomb feature in robotic manufacturing of honeycomb sandwich structure in aerospace field, a localization method using binocular vision to guide industrial robots in single hole filling of honeycomb sandwich structure is proposed. Firstly, the binocular image is segmented into target regions, and the three-dimensional depth map is generated by binocular matching calculation. Then, edge detection, dilation, erosion, skeleton extraction and polygon fitting are carried out on the left camera image to obtain the two-dimensional coordinates of the honeycomb corner feature, and then the actual three-dimensional coordinates are calculated according to the two-dimensional positioning results in depth maps. The position and rotation information of honeycomb holes are obtained by the three-dimensional coordinates. Finally, the positioning experiment of honeycomb structure parts is carried out on the positioning system composed of industrial robot and binocular camera. The experimental results show that the proposed location method can accurately locate honeycomb holes (non-standard hexagon) with certain deformation. The positioning accuracy is less than 2mm, and the positioning accuracy can meet the accuracy requirements of subsequent perfusion operations. |
Key words: industrial robot;machine vision;robot localization;honeycomb structure |