摘要
通过视觉引导机器人完成抓取任务,机器人手眼标定的精度直接影响了抓取任务作业精度和抓取成功率。对于基于位置的机器人视觉引导系统,手眼标定的任务则是确定机器人坐标系与相机坐标系之间的位姿关系。通过HALCON平台,使用线性标定法实现了6DOF机器人的手眼标定。对手眼标定的结果进行反演,直观地展示了手眼标定的精确程度。最后通过采集多组不同数量的图片,在HALCON平台下验证了不同摄像机模型对手眼标定的精度影响,以及同种摄像机模型在不同数量图片的情况下手眼标定的标定精度。实验证明,根据位姿矩阵中待求解的未知量个数采集合适数量的图片和使用更精确的摄像机模型能够提高手眼标定的精度。
Robot hand-eye calibration precision directly influences the precision and the success rate of grasping task. For the position-based visual servo, the task of hand-eye calibration is to determine the relationship about robot coordinate system and camera coordinate system. This paper uses linear calibration method to achieve the 6 DOF robot hand-eye calibration based on the HALCON. It intuitivily displays the hand- eye calibration accuracy by inversing the results of hand-eye calibration. By collecting several groups of different number of images to verify the accuracy of hand-eye calibration under the condition of different camera model and the same model under the condition of different number of images. Experimental results show that according to the number of pose matrix to solve the unknown quantity to collect appropriate quantity of calibration chart and useing more accurate model can improve the accuracy of hand-eye calibration.
出处
《信息技术与网络安全》
2018年第1期97-100,共4页
Information Technology and Network Security
