Omnidirectional humanoid navigation in cluttered environments based on optical flow information

Abstract

In this paper, we address the problem of humanoid navigation in a priori unknown environments, cluttered by obstacles. The robot task is to move within the environment without colliding with obstacles and using only ordinary onboard sensors, like monocular cameras and encoders. The proposed approach relies on: (i) optical flow information, to construct a local representation of the environment obstacles and free space; (ii) visual servoing techniques, to achieve safe motion within the environment while regulating appropriate visual features and the robot internal configuration. In case of navigation in a straight corridor, it can be formally proved that the robot converges to the corridor bisector. With respect to previous works, the algorithm proposed here does not make use of any information about the environment, and exploits the humanoid omnidirectional walking capability to achieve safe navigation in narrow passages. The approach is validated through simulations and experiments with NAO.

Vision-Based Navigation of Omnidirectional Mobile Robots

“…With respect to [9], in this paper we show that the gazevelocity alignment, enforced in [9], represents a steady state constraint in pure visual servoing control schemes that suit the considered robot kinematics and sensors equipment.…”

“…Through a formal comparison of these IBVS schemes we show that a single, appropriate feature and gaze-velocity alignment, as proposed in [9], not only is sufficient to ensure convergence of the robot to the bisector of a corridor but also ensures better transient behavoiur.…”

“…The paper is organised as follows: Section II formulates the problem and summarizes the approach proposed in [9]. Section III presents the considered vision-based strategies and discusses their ability to cope with velocity uncertainties.…”

“…Therefore, we place one/two fictitious centroid/s on the image border of the centroid-free half/ves, to keep x m and y m well-defined. While this alters the true visual features dynamics, it provides a conservative behavior when obstacles are out the camera field of view, preserving the effectiveness of the navigation both in structured scenes (e.g., 90-degreeturns, T-junctions, see [9]), and in more general unstructured cluttered scenarios. The dynamics of the visual features (1) is a function of v r and u p :…”

“…where J and J ωz , are the image Jacobians with respect to the considered velocity inputs [9]. The idea is to regulate to zero the visual features while walking.…”