用于分拣机器人的改进标定方法

A Simple Calibration Method for Sorting Robots

标定使相机的像素坐标关系能转化为有应用意义的实际坐标关系,针对传统张正友标定法在分拣机器人视觉系统应用场景中不能直接建立像素坐标系与机器人坐标系坐标关系的不足,提出一种基于直接线性标定法的改进型简便标定方法,该方法可快速建立机器人坐标系与图像坐标系之间的联系,完成机器人手眼标定。为了验证采用该标定方法的分拣机器人视觉系统所能到达的实际精度,采用重复定位精度0.02mm的SCARA机器人、300万像素相机在120×90mm的视野范围下进行抓取实验,从而验证分拣系统的实际效果。结果表明:所提的算法在实际生产任务环境下可实现0.3mm的定位抓取精度。

Camera coordinate system can be transformed into the practical coordinate relationship of the world, but the camera coordinate system in traditional Zhang method cant convert to robot coordinate directly. In order to solve this problem, it is proposed a method that based on the direct linear calibration method. The method can quickly establish the connection between the robot coordinate system and the image coordinate system, and complete the robot hand - eye calibration. In order to verify the actual accuracy of the sorting robot vision system using the calibration method, a SCARA robot with a repeat positioning accuracy of 0. 02 mm was used, and a 3 million pixel camera was subjected to a crawling experiment in a field of view of 120 x 90 mm to verify sorting. The actual effect of the experiments of systems showed that the method can be used in the sorting robot vision system, can achieve 0. 3mm positioning accuracy, to meet the general needs.