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

Authors:
In Kyu Park;Marcel Germann;Michael D. Breitenstein;Hanspeter Pfister
Affiliations:
Inha University, School of Information and Communication Engineering, 402-751, Incheon, Korea;Swiss Federal Institute of Technology (ETH), Computer Graphics Lab., 8092, Zurich, Switzerland;Swiss Federal Institute of Technology (ETH), Computer Vision Lab., 8092, Zurich, Switzerland;Harvard University, School of Engineering and Applied Sciences, 02138, Cambridge, MA, USA
Venue:
Machine Vision and Applications - Integrated Imaging and Vision Techniques for Industrial Inspection
Year:
2010

Citing 17
Cited 0

A Method for Registration of 3-D Shapes

IEEE Transactions on Pattern Analysis and Machine Intelligence - Special issue on interpretation of 3-D scenes—part II
Object modelling by registration of multiple range images

Image and Vision Computing - Special issue: range image understanding
Numerical recipes in C (2nd ed.): the art of scientific computing

Numerical recipes in C (2nd ed.): the art of scientific computing
COSMOS-A Representation Scheme for 3D Free-Form Objects

IEEE Transactions on Pattern Analysis and Machine Intelligence
Multidimensional Morphable Models: A Framework for Representing and Matching Object Classes

International Journal of Computer Vision
Using Spin Images for Efficient Object Recognition in Cluttered 3D Scenes

IEEE Transactions on Pattern Analysis and Machine Intelligence
A morphable model for the synthesis of 3D faces

Proceedings of the 26th annual conference on Computer graphics and interactive techniques
Vision guided bin picking and mounting in a flexible assembly cell

IEA/AIE '00 Proceedings of the 13th international conference on Industrial and engineering applications of artificial intelligence and expert systems: Intelligent problem solving: methodologies and approaches
Real-time 3D model acquisition

Proceedings of the 29th annual conference on Computer graphics and interactive techniques
Geometric Probing of Dense Range Data

IEEE Transactions on Pattern Analysis and Machine Intelligence
Combining greyvalue invariants with local constraints for object recognition

CVPR '96 Proceedings of the 1996 Conference on Computer Vision and Pattern Recognition (CVPR '96)
Discrete Pose Space Estimation to Improve ICP-Based Tracking

3DIM '05 Proceedings of the Fifth International Conference on 3-D Digital Imaging and Modeling
Jump flooding in GPU with applications to Voronoi diagram and distance transform

I3D '06 Proceedings of the 2006 symposium on Interactive 3D graphics and games
Three-Dimensional Model-Based Object Recognition and Segmentation in Cluttered Scenes

IEEE Transactions on Pattern Analysis and Machine Intelligence
Robust global registration

SGP '05 Proceedings of the third Eurographics symposium on Geometry processing
Automatic Pose Estimation for Range Images on the GPU

3DIM '07 Proceedings of the Sixth International Conference on 3-D Digital Imaging and Modeling
Finding optimal views for 3D face shape modeling

FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition

Quantified Score

Hi-index	0.00

Visualization

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.