面向机器人激光增材制造的机器视觉系统标定算法

随着现代智能制造的快速发展,金属增材制造、绿色再制造、焊接等领域大量使用机器人等自动化装备,视觉传感是机器人智能制造的关键环节。针对上述需求设计了一套基于线结构光的视觉传感器,并对其进行标定,自主开发了用于机器人激光增材制造的机器视觉系统。该系统通过Matlab相机标定工具包实现相机的内外参数标定,利用Labview编程提取出像素坐标,根据最小二乘法拟合出相机坐标系下的光平面方程,实现像素坐标到光平面坐标的转换,最后针对视觉传感器与机器人的"Eye-in-Hand"系统实现手眼标定,完成像素坐标到机器人三维基坐标的转换。实验结果表明,开发的视觉传感器具有较高的定位精度,各方向平均误差仅为1 mm。

Laser Additive Manufacturing Machine Vision System Development

With the rapid development of modern intelligent manufacturing, Robots and other automation equipment are widely used in metal additive manufacturing, green remanufacturing, welding and other fields. Visual sensing is a key link in robot intelligent manufacturing. This article addresses the above requirements, Designed and calibrated a set of vision sensors based on laser structured light, and independently developed a machine vision system for the robotic laser additive manufacturing process. This method realizes the internal and external parameter calibration of the camera through the Matlab camera calibration toolkit, uses Labview programming to extract the pixel coordinates, fits the laser plane equation in the camera coordinate system according to the least square method, and realizes the conversion from pixel coordinates to laser plane coordinates. Finally, the "Eye-in-Hand" system of the vision sensor and the robot realizes hand-eye calibration, and obtains the conversion of pixel coordinates to the three-dimensional base coordinates of the robot. Experiments show that the developed vision sensor has high positioning accuracy.