Moving To Oop: An Active Observation Approach For A Novel Composite Visual Servoing Configuration

The imaging quality of a camera has an important influence on the results of visual servoing. To improve the imaging quality, a novel active-observation-based composite visual servoing ( AOBCVS) combined with eye-in-hand and eye-to-hand configuration can be adopted, which consists of a mobile observer and a manipulator. One main problem of this configuration that should be solved is to endow the mobile observer with active observation ability. In order to obtain the optimal observation effect of the two markers (one marker can be seen as a target, and the other marker can be attached to a manipulator), a two-level scheme is designed for a mobile robot; in the decision level, a vehicle can determine the optimal observation position (OOP) by optimizing an observation measure function (OMF), and in the control level, the vehicle is driven to the OOP and a pantilt camera mounted on it keeps the two markers in the view. The scheme of observer has such advantages: 1) a satisfactory solution for the field-of-view problem; 2) a comprehensive OMF; and 3) an ability to adjust the camera position in real time based on the current observation effect. Finally, we provide the simulation and experimental results to validate the effectiveness of the composite configuration.