Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Real-Time Visual Feature Recognition and Metrology for Mechanical Micro-Parts
Machine vision has lately been gaining in importance in various fields. One possible application lies in robot-assisted micro-assembly. Your task will be to devise an algorithm that recognizes edges and features of micro-parts, and intuitively displays the results to the user to support him.
In the last decade, machine vision has been tremendously gaining in importance in various fields, ranging
from robotics and autonomous systems to healthcare, agriculture, and consumer electronics. A key aspect enabling
this rapid development - particularly for the consumer market - is the massive price drop for optical cameras due to
their fast development by the telecommunication industry. One fundamental and recurring task in machine vision
applications is the detection of edges in an image; two examples of standard edge detectors can be seen in the enclosed pdf.
This information can then further be utilized for various purposes, for instance geometric feature recognition or
metrology. One of the applications that can largely benefit from such algorithms is robot-assisted micro-assembly,
where a robotic system is used to assist the user in assembling mechanical micro-parts into a functional micro-device.
Although considerable effort has been put into the engineering of such systems in the last few decades, see for instance the example in the enclosed pdf, the integration of smart machine vision algorithms intuitively supporting the user remains a challenge.
In the last decade, machine vision has been tremendously gaining in importance in various fields, ranging from robotics and autonomous systems to healthcare, agriculture, and consumer electronics. A key aspect enabling this rapid development - particularly for the consumer market - is the massive price drop for optical cameras due to their fast development by the telecommunication industry. One fundamental and recurring task in machine vision applications is the detection of edges in an image; two examples of standard edge detectors can be seen in the enclosed pdf. This information can then further be utilized for various purposes, for instance geometric feature recognition or metrology. One of the applications that can largely benefit from such algorithms is robot-assisted micro-assembly, where a robotic system is used to assist the user in assembling mechanical micro-parts into a functional micro-device. Although considerable effort has been put into the engineering of such systems in the last few decades, see for instance the example in the enclosed pdf, the integration of smart machine vision algorithms intuitively supporting the user remains a challenge.
The key task within a micro-assembly process is the precise insertion of mechanical microparts
into the overall assembly. This also includes the fixation of said parts for instance by the means of miniature
screws. Typical problems arising during this task include poor alignment of parts before insertion and undetected
manufacturing inaccuracies, both leading to jamming or even damaging of the parts during insertion. To remedy
these issues, it can be helpful to provide the user with visual cues about the exact location and orientation of holes
and threads, and about inaccuracies in the part geometry. Your task will be to develop a system that obtains these
information from an optical camera by using appropriate machine vision algorithms, and to display them to the user
in real-time.
The key task within a micro-assembly process is the precise insertion of mechanical microparts into the overall assembly. This also includes the fixation of said parts for instance by the means of miniature screws. Typical problems arising during this task include poor alignment of parts before insertion and undetected manufacturing inaccuracies, both leading to jamming or even damaging of the parts during insertion. To remedy these issues, it can be helpful to provide the user with visual cues about the exact location and orientation of holes and threads, and about inaccuracies in the part geometry. Your task will be to develop a system that obtains these information from an optical camera by using appropriate machine vision algorithms, and to display them to the user in real-time.