Vision Calibration for Industrial Robots An Approach using Cognex VisionPro

06/02/2017

Most manufacturing processes rely upon the use of industrial robot arms for different applications, such as inline robotics, system assembly robotics and machine tending robotics. In the pursuit of the highest industrial quality standards, accuracy, precision, flexibility and a larger potential to increase the competitiveness in the market, some industrial robot cells make use of vision sensors. These vision sensors and cameras are used to guide the robot cells in pick and place, perform measurements and other applications. In order to successfully execute these tasks, there is the need to relate the cameras’ 2D projections with 3D positions of the robot’s coordinate system, which can only be done after a calibration process. This master thesis project’s main focus is to develop a standard application capable of automatically calibrating the cameras in the robot’s workspace, by estimating their exact position and orientation with respect to the robot, while also compensating for the non-linear distortion of its lenses. At the same time, 3D tools are proposed and introduced. The developed calibration procedures not only show an improvement in typical applications up to 11 times, when comparing a calibrated and uncalibrated camera, but are also responsible for significantly decreasing human effort in the process of teaching the tools’ positions to the robot. In conclusion, the mentioned procedures were also implemented with success at the company at which this project has been developed.

Authors in the community:

• 

  M

  Manuel Lemos Ribeiro

  ist173773

Supervisors of this institution:

• 

  L

  Leonel Augusto Pires Seabra de Sousa

  ist13348

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

eng - English

Instituto Superior Técnico