2D model-based tracking of complex shapes for visual servoing tasks

Visual servoing needs image data as input to realize robotics tasks such as positioning, docking or mobile target pursuit. This often requires to track the 2D projection of the object of interest in the image sequence. To increase the versatility of visual servoing, objects cannot be assumed to carry landmarks. We have developed an original method for 2D tracking of complex objects which can be approximately modeled by a polyhedral shape. The proposed method fulfills real-time constraints as well as reliability and robustness requirements. Real experiments and results on a positioning task with respect to different objects are presented.