Identification of part pose

Summary

This component uses the information provided by a 2D vision camera to estimate the pose of a part in the space, i.e., the X, Y, Z, pitch, roll and yaw values of the part with respect to the camera frame.

The configuration of the component requires a training phase in which a sample of the part and its CAD file are used (stl format).

It is implemented as a service in a ROS node, which also includes tools to calibrate the vision system with respect to the robot, so that the pose can be referred to it, and allows using either both external cameras or the cameras mounted on the robot (eye-in-hand configuration).

Optionally, it can be delivered as an executable program in Windows with a tcp/ip interface

Results type(s)
Unfold all
/
Fold all
More information & hyperlinks