No abstract

In this paper we present the results from our visual servoing system for a real industrial robot. In contrast to other visual servoing system in the literature, the one presented here, which was derived from [4], uses a quaternion representation of the rotation instead of the more common matrix representation. By doing so, the proposed system avoids potential singularities introduced by the rotational matrix representation. After performing exhaustive tests in a simulated environment, our controller was applied to a Kawasaki UX150. In the case of simulation, the movement of the camera and the image processing were performed using Matlab-Simulink, which allowed us to test the controller regardeless of the mechanism in which the camera was moved and the underlying controller that was needed for this movement. In the case of the real robot, the controller was tested initially using another simulation program provided by Kawasaki Japan and later with the real Kawasaki robot. The setup for testing and the results for all three cases above are presented here, but for more details on the simulations, the reader is encouraged to check [8].

Particle ltering (also known as the condensation algorithm) has been widely applied to model-based human motion capture. However, the number of particles required for the algorithm to work increases exponentially with the dimensionality of the model. In order to alleviate this computational explosion, we propose a two-level hierarchical framework. At the coarse level, the conguration space is discretized into large partitions and a suboptimal estimation is calculated. At the ne level, new particles in the vicinity of the suboptimal estimation are created using a more likely and narrow conguration space, allowing the original coarse estimate to be rened more eciently.Our preliminary results demonstrates that this hierarchical framework achieves accurate estimation of the human posture with signicantly reduction in the number of particles.Keywords: coarse-to-ne, bottom-up aggregation of state estimations. we consider a model-based 3D human motion capture using particle lters.Particle ltering has proved to be an eective and robust technique for contour tracking [13,14,15].However, the application of particle ltering to mark- In this work, we address the problem of dimensionality and the associated computational complexity byproposing an ecient coarse-to-ne framework using monocular image sequences. The rest of this paper

No abstract

This paper presents a completely autonomous camera calibration framework for a vision sensor network consisting of a large number of arbitrarily arranged cameras. In the proposed framework, a sequence of images for calibration is collected without a tedious human intervention. Next, the system automatically extracts all necessary features from the images and finds the best set of images that minimizes the error in 3D reconstruction considering all cameras in the set.