Machine Vision and Applications

, Volume 21, Issue 5, pp 749–766

Fast and automatic object pose estimation for range images on the GPU

  • In Kyu Park
  • Marcel Germann
  • Michael D. Breitenstein
  • Hanspeter Pfister
Original Paper

Abstract

We present a pose estimation method for rigid objects from single range images. Using 3D models of the objects, many pose hypotheses are compared in a data-parallel version of the downhill simplex algorithm with an image-based error function. The pose hypothesis with the lowest error value yields the pose estimation (location and orientation), which is refined using ICP. The algorithm is designed especially for implementation on the GPU. It is completely automatic, fast, robust to occlusion and cluttered scenes, and scales with the number of different object types. We apply the system to bin picking, and evaluate it on cluttered scenes. Comprehensive experiments on challenging synthetic and real-world data demonstrate the effectiveness of our method.

Keywords

Object pose estimation Bin picking Range image processing General purpose GPU programming Iterative closest point Euclidean distance transform Downhill simplex CUDA 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag 2009

Authors and Affiliations

  • In Kyu Park
    • 1
  • Marcel Germann
    • 2
  • Michael D. Breitenstein
    • 3
  • Hanspeter Pfister
    • 4
  1. 1.School of Information and Communication EngineeringInha UniversityIncheonKorea
  2. 2.Computer Graphics Lab.Swiss Federal Institute of Technology (ETH)ZurichSwitzerland
  3. 3.Computer Vision Lab.Swiss Federal Institute of Technology (ETH)ZurichSwitzerland
  4. 4.School of Engineering and Applied SciencesHarvard UniversityCambridgeUSA

Personalised recommendations