Modern robots and mechatronic systems are equipped with a large number of sensors which give them the possibility to accomplish complex tasks. In recent years the integration of robots and vision system has become critical to improve machinery productivity and versatility. Nevertheless, the most widely used approach is still the "look then move" fashion, in which the vision system simply supports a manipulator without taking part in the robot control task. An alternative technique that allows to merge the vision task and the control one is named "visual servoing" or "vision in the loop". Following this approach, a visual-feedback loop can be added to the standard manipulator control loop in order to drive the robot directly with the information coming from the vision system. This further controller brings a lot of benefits including: (i) a general increase of the accuracy of the subsystem and its robustness towards environment noise and changes; (ii) the capability to follow fast moving objects; (iii) the possibility to adapt the system to different situations and tasks. On the contrary hand, designing and developing an efficient and robust visual control is tipically a difficult task as it involves various issues like computer vision lgorithms, control theory, robot kinematics and so on. The analysis carried out in this thesis aims to highlight the aspects common to all visual servoing tasks in order to provide a general solution schema for this kind of problems. Firstly, the image acquisition process and image elaboration techniques are presented, with particular attention to their application to "tracking" fast moving objects or scenes; secondly, visual servoing controller design is ntroduced and its kinematic and dynamic issues are studied; eventually, a software architecture for the mplementation of visually guided task is shown along with simple experimental applications used to evaluate the results of the work.
Un approccio meccatronico per la progettazione e lo sviluppo di sistemi di controllo in asservimento visivo
ROSSETTI, MATTIA
2015
Abstract
Modern robots and mechatronic systems are equipped with a large number of sensors which give them the possibility to accomplish complex tasks. In recent years the integration of robots and vision system has become critical to improve machinery productivity and versatility. Nevertheless, the most widely used approach is still the "look then move" fashion, in which the vision system simply supports a manipulator without taking part in the robot control task. An alternative technique that allows to merge the vision task and the control one is named "visual servoing" or "vision in the loop". Following this approach, a visual-feedback loop can be added to the standard manipulator control loop in order to drive the robot directly with the information coming from the vision system. This further controller brings a lot of benefits including: (i) a general increase of the accuracy of the subsystem and its robustness towards environment noise and changes; (ii) the capability to follow fast moving objects; (iii) the possibility to adapt the system to different situations and tasks. On the contrary hand, designing and developing an efficient and robust visual control is tipically a difficult task as it involves various issues like computer vision lgorithms, control theory, robot kinematics and so on. The analysis carried out in this thesis aims to highlight the aspects common to all visual servoing tasks in order to provide a general solution schema for this kind of problems. Firstly, the image acquisition process and image elaboration techniques are presented, with particular attention to their application to "tracking" fast moving objects or scenes; secondly, visual servoing controller design is ntroduced and its kinematic and dynamic issues are studied; eventually, a software architecture for the mplementation of visually guided task is shown along with simple experimental applications used to evaluate the results of the work.File | Dimensione | Formato | |
---|---|---|---|
Rossetti Mattia - PHD Thesis.pdf
accesso aperto
Dimensione
23.13 MB
Formato
Adobe PDF
|
23.13 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/67591
URN:NBN:IT:UNIBG-67591