基于机器人视觉伺服的确定下一最优视点的方法研究

Studies to determine the next best view of the robot visual servo- based approach

机器人自动视点规划是目标物自动三维重构研究中的首要问题,而自动确定下一最优视点的位置和方向是自动视点规划的关键。经典群矢量链法只能确定下一最优视点的方向,该文提出一种简化的2-1/2-D视觉伺服新方法,确定下一最优视点的空间位置。新方法使用群矢量链法以开环的方式控制末端执行器的方向;通过图像质心坐标与目标物图像面积等图像特征,采用基于图像的视觉伺服控制末端执行器的位置。方向控制与位置控制相结合形成2-1/2-D的控制策略,使机器人达到下一最优视点的期望位姿。新方法图像雅可比矩阵维数低,控制方法相比经典2-1/2-D视觉伺服更简单易实现。实验结果显示,该方法可使机器人运动到期望的方向和位置实现自动视点规划。最终生成的目标物曲线表面基本上是完整的,只有因目标物表面有些凹小区域引起的很小盲区。

In automatic 3D reconstruction, automatic view planning is a primary task, where deter- mining next best view （NBV） is the key problem. A visual servo - based approach is proposed in this pa- per for determining the NBV spatial position, where the 2 - 1/2 - D visual servoing is involved. The clas- sical mass vector chain （MVC） is used for determining the NBV direction. The orientation of the end ef- fector is controlled by MVC method in an open - loop, which decreases the dimensionality of the image Jacobian. Some image features are extracted for position control of the end effector in 2 - 1/2 - D visual servoing. The experimental results show that the proposed method can guide the robot to the desired ori- entation and position for automatic multi - view planning. The final produced curve surface is basically complete, only with a few tiny blind areas possibly because of many concave small surface patches on sur- face.