Vision‐based Control

Gangloff, Jacques; Nageotte, Florent; Poignet, Philippe

doi:10.1002/9781118562147.ch6

Cited by 1 publication

(1 citation statement)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…), visual servoing is applied for grasping and placing tasks (Wu et al (2017)). In medicine, visual information from endoscopies or Computed Tomography (CT) scans is used to manoeuvre robot manipulators during surgeries (Gangloff, Nageotte, and Poignet (2013); Azizian et al (2014Azizian et al ( ), (2015). In the manufacturing industry, visual servoing is used in many applications such as welding, assembly, painting, etc.…”

Section: Introductionmentioning

confidence: 99%

Generalization of reference filtering control strategy for 2D/3D visual feedback control of industrial robot manipulators

Solanes

Muñoz-Benavent

Armesto

et al. 2021

International Journal of Computer Integrated Manufacturing

View full text Add to dashboard Cite

This paper develops the application of the Dual Rate Dual Sampling Reference Filtering Control Strategy to 2D and 3D visual feedback control. This strategy allows to overcome the problem of sensor latency and to address the problem of control task failure due to visual features leaving the camera field of view. In particular, a Dual Rate Kalman Filter is used to generate inter-sample estimations of the visual features to deal with the problem of vision sensor latency, whereas a Dual Rate Extended Kalman Filter Smoother is used to generate more convenient visual features trajectories in the image plane. Both 2D and 3D visual feedback control approaches are widely analyzed throughout the paper, as well as the overall system performance using different visual feedback controllers, providing a set of results that highlight the improvements in terms of solution reachability, robustness, and time domain response. The proposed control strategy has been validated on an industrial system with hard real-time limitations, consisting of a 6 DOF industrial manipulator, a 5 MP camera, and a PLC as controller.

show abstract