| In this paper, the main researches aim at robotic visual servoing control, four servoing control methods in common use are introduced in detail, namely position-based servoing control, image-based servoing control, 2.5 D servoing control and visual servoing based on active contour. By using pose error of camera between current pose and desired pose, the visual servoing method of position-based designs controller in robotic task space, the main advantage of this approach is that it directly controls the camera trajectory in Cartesian space. A resolved acceleration PID visual servoing controller using RPY angles is designed, the camera can track or position object with accelerated motion. The error signal in image-based visual is direct defined in image space, which insures object in camera field of view all along. There are two image-based servoing control presented, one is optimal control law of robot based on delta operator in visual servoing. There is very rare in designing optimal control law using delta transform. Another is adaptive servoing visual robot control based on moment. The image Jacobian matrix is structured by image moment, which overcomes weak match precision dues to texture, illumination, noises. The method compensates dynamics equation also. Combining position-based and image-based visual servoing, then presents 2.5D visual servoing. The error to be minimized is specified both in the image and in the pose space. Comparing to the position-based visual servoing, the 2 1/2 D control scheme does not need any geometric 3D model of the object. Furthermore and comparing to image-based visual servoing, the approach ensures the convergence of the control law in the whole. The robustness of the visual servoing scheme with respect to camera calibration errors is well. A fourth visual servoing based on active contour is also introduced in the paper, which allows the tracking of planar objects in the environment, and a low-cost extraction of visual parameters from raw image data.
|
PDF Full Text Request