home   |   research   |   teaching   |   publications   |   biosketch   |   CV

Image-based Visual Servoing for Mobile Robot
Using Epipolar Geometry

We present an image-based visual servoing strategy for driving a nonholonomic mobile robot equipped with a pinhole camera (see Fig. 1(left)) toward a desired configuration. The proposed approach uses the epipolar geometry (see Fig. 1(right)) and does not need any knowledge of the 3-D scene geometry. Asymptotic convergence to the desired configuration is proven, both in the calibrated and partially calibrated case.
Figure 1: (Left) PIONEER robot with pinhole camera. (Right) The epipolar geometry is used as the main tool in our servoing strategy.
A generical starting configuration is shown in Fig. 2(left), while the corresponding position of feature points and epipoles (here superimposed to the desired image) is shown in Fig. 2(right). The control scheme is divided in two steps. In the first, using an approximate input-output linearizing feedback, the epipoles are zeroed, so as to align the robot with the goal. Feature points are then used in the second translational step to reach the desired configuration. Global asymptotic convergence has been proven, both in the calibrated and partially calibrated case.
Figure 2: (Left) The IBVS strategy based on epipolar geometry drives the robot form initial to target position (in which the desired view is known) using the current view. (Right) From the pinhole camera we extract desired (circles) and current (crosses) features to compute epipoles (ed and ea) (here superimposed to the desired image) and control the robot.


We present here results from the experimental session:



We wish to thanks Nicola Pisu and Riccardo Donati for their invaluable help in the realization of experimental results.

File translated from TEX by TTH, version 3.85.
On 4 Jan 2009, 22:46.