In this paper the assembly of micro-optic components using image processing is introduced. An automated positioning process for laser resonators, which is handled manually until now, is implemented in a prototypical scale. For this work a cartesian precision-robot is used in combination with an image processing system including two cameras with different resolutions and perspectives. The components to be positioned are presented orderless on a pallet and have to he detected, gripped and aligned with an edge after calculating position and angle correction values. With a camera that is mounted vertically on the Z-axis of the robot the position of the components on the pallet is detected. Although vacuum grippers are typically not able to centre components one has to be used in this process as the effecting surface on the sides of the laser resonators may not be touched. To correct the errors that were caused by the suction of the vacuum gripper a second image processing is used to get values for a correction of position and angle. These images are taken with the help of a mirror that is positioned in an angle of 450 With this mirror a horizontal mounted camera can take pictures of the gripper holding the laser resonator from underneath. Very good results could be achieved during an experimental evaluation, caused by the second image processing in particular. The results showed that it is possible to get a position error of only a few micrometers. In the whole process primarily the lighting system was identified as the critical factor.
展开▼
