Improving visually servoed disassembly operations by automatic camera placement

Abstract

It is advantageous to use an independently mobile camera in visually servoed disassembly operations since its degrees of freedom can be used to bring the camera into an optimal view-pose. Apart from the vision-based closed-loop control that guides the disassembly robot, a second vision-based closed-loop control can be set up which upgrades the camera from a passive to an active sensor and optimizes the view-pose according to changes in the scene domain. The objective of this contribution is to present an approach to model-based pose initialization and pose refinement steps of an active camera which is able to simultaneously observe and track several objects moving to some extent independently from each other. In both steps, constraints regarding object occlusions, the limited field-of-view, and the uncertainty to pose estimation are exploited to determine an optimal camera pose. The feasibility of this approach to model-based pose initialization is experimentally validated and document ed for the case of the manipulation of a shock absorber lid

3D Sensor planning framework for leaf probing

“…These approaches are often based on visual servoing techniques [6]. Here, planning algorithms find optimal solutions and consequently, execute the minimum number of views to complete the task.…”

Optimal arrangement of ceiling cameras for home service robots using genetic algorithms

“…Past research on camera placement was mostly limited to the placement of only one sensor [2,3], two sensors [4] or to the special case of marker detection [5]. The problem of camera placement is closely related to the guard placement Manuscript problem (AGP).…”

Visually Guided Cooperative Robot Actions Based on Information Quality

“…Researchers have used model-based approaches, requiring previously known environments (Burschka, 1997;Cowan, 1988;Hutchinson, 1989;Kececi, 1998;Laugier, 1995). Target motions (if any) are assumed to be known (Laugier, 1995).…”

“…Target motions (if any) are assumed to be known (Laugier, 1995). Brute force search methods divide the view volume (into grids, octrees, or constraint sets), and search algorithms for optimum sensor location are applied (Connolly, 1985;Cowan, 1988;Kececi, 1998;Luo, 1989). These methods require a priori knowledge of object/target models (Chu, 1997;Tarabanis, 1995).…”