Position based visual servoing: keeping the object in the field of vision

Abstract

Visual servoing requires an object in the field of view of the camera, in order to control the robot evolution. Otherwise, the virtual l i i is broken and the control loop ,cannot continue to be closed.In this paper, a novel approach is presented in order to guarantee that the object remains in the field of view of the camera during the whole robot motion. It consists in t r d n g an iteratively computed trajectory. A position based modeling adapted to a moving target object is established, and is used to control the trajectory. A nonlinear decoupliig approach is then used to control the robot. Experiments, demonstrating the capabilities of thiS a p proach, have been conducted on a Cartesian robot connected to a real time vision system, with a CCD camera mounted on the end effector of the robot.

Optimizing plane-to-plane positioning tasks by image-based visual servoing and structured light

“…Then, noting , we obtain It is obvious that simple combinations of such features can lead to a decoupled system. To do that, we have chosen the following set of visual features: (11) Note that , , or are never equal to zero, except in the degenerate case when the camera is at infinity. The interaction matrix of is (12) which is always of rank 3 and never singular.…”

“…Let us note that if a simple triangulation algorithm is used to estimate the 3-D parameters , , and , the results are worse, since a very unstable behavior is obtained (not shown here). Furthermore, another drawback of this technique is that, like in all position-based approaches, the stability cannot be analytically proven, as it depends on the convergence of the reconstruction algorithm [11].…”

“…A suitable design strategy attempts to decouple visual features, so that each one is closely related to one degree of freedom (DOF). Of course, position-based visual servoing provides such a decoupling [11], [12], but the problem with such methods is with the stability of the pose-estimation algorithm with respect to image noise [13]. Hybrid techniques are based on controlling rotational DOFs in the Cartesian space while the translational ones are controlled from image data [9], [14].…”

A Pursuit-Rendezvous Approach for Robotic Tracking

“…An algorithm for tracking humans from a moving platform is suggested in (Davis et al, 2000). Visual servoing methods are also used to keep the target in the field of vision of the robot (Thuilot et al, 2002). This problem is related to the problem of positioning and localization of the robot with respect to the moving object (Kim et al, 2001;Chaumette et al, 1999).…”

Visual Servoing: Reaching the Desired Location Following a Straight Line via Polynomial Parameterizations

“…Many methods for dealing with such a task have been proposed. Some examples are the position-based visual servoing (PBVS) where the feedback error is the camera pose (see for example [4], [5]), and the image-based visual servoing (IBVS) where the feedback error is the image error (see for example [6], [7]). In the 2 1/2 D visual servoing [8] the feedback error contains both the camera pose and the image error.…”