Binocular stereo vision based humanoid manipulator control

Inverse kinematics of manipulator and object location are the key technology for service robot. In this paper, a binocular stereo vision based humanoid manipulator control approach is presented for the service robot which is working in the “human friendly” environments. An inverse kinematic solver is proposed to find all joint angles for given position of the effectors on the manipulator. The target object is recognized by color segmentation. The 3D position is computed by the stereo vision system after contour extraction. Finally, the approach is applied to the service robot with Bumblebee2 camera and two arms for object grasping. The experimental results are conducted for illustration of effectiveness of the proposed methods.