Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
A Simplified Pose Estimation Algorithm for Bin Picking: Using the convex hull of the CAD model
University West, Department of Engineering Science, Division of Production System.
2017 (English)Independent thesis Advanced level (degree of Master (One Year)), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

The applications of robotics are becoming more and more popular. Lots of tasks in the industry like welding, painting, handling, assembly, casting and so on are handled by robots now. Currently most industrial robots on the production line are controlled by pre-teaching which is not effective and flexible. The automation of robots has been researched since the first robot introduced to the industry. The basic task like the automation of picking objects for a robot is still a challenge. University west wants to find a faster automation picking method for industrial robots. This thesis presents a simplified pose estimation algorithm for bin picking by using the CAD model of the object. The main idea is to simplify the pose estimation task by identifying all stable positions of the object and predefining a picking point for each stable position according to the data of its CAD model in advance. Then the online work focuses only on the image analysis in 2D which is simple so as to achieve a fast picking. The experimental results satisfy the requirements. First, all positions of the object are found by checking the convex hull of its CAD model. A stable position is identified as having the centre of gravity above the convex hull of the support surface of the position. Then virtual images are generated using a computed virtual camera, having the same parameters as the calibrated camera, in all stable positions of the object, seen from above. All virtual images are classified into different classes for the preparation of the online classification. Picking points for each stable position in the virtual images are predefined. The 3D data of each picking point is calculated according to the data of its CAD model in advance. Finally, the online work finds the class which the real image represents and the predefined picking point. The final pose is estimated by the position transformed from the image coordinate system to the world coordinate system using the camera position and rotation. This algorithm focuses on well-defined objects having a limited number of stable positions and only non-zero area support surfaces. Further works will be the trajectory planning which should avoid conflict by using the CAD model of the object.

Place, publisher, year, edition, pages
2017. , p. 17
Keywords [en]
Bin picking, pose estimation, CAD model, convex hull
National Category
Production Engineering, Human Work Science and Ergonomics
Identifiers
URN: urn:nbn:se:hv:diva-12026Local ID: EXM810OAI: oai:DiVA.org:hv-12026DiVA, id: diva2:1178219
Subject / course
Robotics
Educational program
Robotteknik
Supervisors
Examiners
Available from: 2018-01-29 Created: 2018-01-29 Last updated: 2018-01-29Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
Division of Production System
Production Engineering, Human Work Science and Ergonomics

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 31 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf