Thesis - Registration in the application of bin picking

Point cloud registration, finding an unknown transform between point clouds, is often researched in relation to motion estimation or SLAM. But it is also useful in object detection and pose estimation. One industry application is random bin picking. That is picking known randomly oriented objects from a container. Some differences in requirements compared to e.g. SLAM applications are:

• There are multiple instances of the matched object. The instances are usually very close to each other in space.
• Due to the static scene the noise of the 2.5D sensor might be predictable. At least more so than the noise from a car mounted or handheld sensor.
• The objects often occlude each other. Leading to less overlap between the point clouds.
  However, since there is usually a 3D model of the object, only one of the point clouds is partial.
• Due to the simplicity of some of the picked objects. There are often a symmetry line or multiple identical feature points in the same instance of the object.

The task for this thesis will be to investigate the following:

• Does any of these differences make a certain state of the art method more/less suitable for bin picking?
• Are there any improvements that can be made due to these limitations/assumptions?

Prerequisites

This thesis will require basic knowledge in computer vision, machine learning and deep learning. There will also be python programming and data wrangling involved.

Contact

For more information about the position, contact:

Christoffer Malmgren, Algorithm developer, christoffer.malmgren@sick.se

or

Charlotte Axelsson, HR Manager, charlotte.axelsson@sick.se

Welcome with your application 14^th of October at the latest!

SICK is a world-leading supplier of sensors and sensor solutions for industrial applications. We are 11 000 employees in 50 countries and our headquarter is located in Freiburg, Germany. SICK in Linköping is an innovation center for Machine Vision and we are 80 committed employees with a big interest in image processing and visualization. For more than 35 years, our team at SICK Linköping has successfully developed and delivered software for technically leading products within the field of 2D and 3D vision, as well as system solutions for i.e. robot guidance and quality control.

At SICK in Linköping, we are very proud of being a healthy and attractive workplace. For many years, we have been elected as one of the best workplaces in Sweden according to the survey Great Place to Work, the latest award is from 2021. We work actively to reduce our climate footprint and we are active in various ways to contribute to the society and to increase diversity at our workplace.