Project-Homepage

Project: 2D-3D Pose Estimation

Researchers: Rosenhahn B. , Perwass C. , Granert O. , Sommer G.

This research project deals with the 2D-3D pose estimation problem. Pose estimation itself means to estimate the relative position and orientation (rotation and translation) of a (known) 3D object with respect to a reference camera system.

The main focus of our work concentrates on the geometric modeling and application of the pose problem. Therefore we use as mathematical language the conformal geometric algebra, which is a universal Clifford algebra. It provides a homogeneous model for stereographically projected points on a hypersphere and therefore couples kinematics with projective geometry. This representation is useful for many computer vision or robot vision tasks.
For the pose scenario, 3D object models are treated two-fold, feature-based and free-form based:

The reason for this approach is that starting from simple features (e.g. point features) an extended scenario is derived, which deals with higher order features such as lines, planes, circles, spheres, kinematic chains or cycloidal curves. Cycloidal curves are defined as circles rolling on circles and are treated as 3D entities within the pose scenario. This scenario extends to general free-form contours by interpreting contours generated with 3D Fourier descriptors as n-times nested cycloidal curves.
The following movies visualize some properties of our developed algorithms (click on the images):

Our algorithm for pose estimation of kinematic chains is robust and can be used for different tasks, e.g. for a visual remote controling of a manipulator. PoseHouse.gif Solve000.gif
In our free-form pose estimation algorithms we are able to detect outliers during pose estimation and additionally to deal with multiple and partial hidden object contours. BoneStab SaumTeller
The next sequence shows pose results of two nested contours. The other sequence shows further the possibility to model slight object deformations by fusing kinematic chains within our approach for free-form object modeling. Tree Morph
Our future work will continue with free-form pose estimation of surfaces.
The developed algorithms are also applied in our navigation project for selflocalization of mobile robots.

Questions, remarks:

Bodo Rosenhahn
Christian Perwass
Oliver Granert
Gerald Sommer

Publications:

2004 Pose Estimation of Free-form Objects
Rosenhahn B., Sommer G., Klette R.
Technical Report 0401, Christian-Albrechts-Universität zu Kiel, Institut für Informatik und Praktische Mathematik, März 2004
PDF, Bibtex
2003 Pose estimation of free-form surface models
Rosenhahn B., Perwass C., Sommer G.
In 25. Symposium für Mustererkennung, DAGM 2003, Magdeburg, Vol. 2781 of LNCS, pp. 574-581, Springer-Verlag, Berlin, 2003.
PDF, Bibtex
2003 Pose Estimation Revisited
Rosenhahn B.
Dissertation, Institut für Informatik und Praktische Mathematik, Christian-Albrechts-Universität zu Kiel, 2003.
PDF, Bibtex, Abtract
2002 Pose Estimation of 3D Free-form Contours
Rosenhahn B., Perwass C., Sommer G.
Technical Report 0207, Christian-Albrechts-Universität zu Kiel, Institut für Informatik und Praktische Mathematik, August 200 2
PDF, Bibtex
2002 Pose Estimation in Conformal Geometric Algebra, Part I: The Stratification of Mathematical Spaces, PartII: Real-time Pose Estimation us ing Extended Feature Concepts
Rosenhahn B., Sommer G.
Technical Report 0206, Christian-Albrechts-Universität zu Kiel, Institut für Informatik und Praktische Mathematik, November 2 002
PDF, Bibtex
2002 Pose estimation of 3D free-form contours in conformal geometry
Rosenhahn B., Perwass C., Sommer G.
In D. Kenwright, editor, Proceedings of Image and Vision Computing, IVCNZ, Auckland, NZ, pp. 29-34. 2002
PDF, Bibtex
2002 Adaptive pose estimation for different corresponding entities
Rosenhahn B., Sommer G.
In L. Van Gool, editor, Pattern Recognition, 24. Symposium für Mustererkennung, Zürich, September 2002, Vol. 2449 of LNCS, pp. 2 65-273. Springer-Verlag, Berlin Heidelberg, 2002
PDF, Bibtex
2002 Monocular pose estimation of kinematic chains
Rosenhahn B., Granert O., Sommer G.
In L. Dorst, C. Doran and J. Lasenby, editors, Applications of Geometric Algebra in Computer Science and Engineering, pp. 373-37 5. Proc. AGACSE 2001, Cambridge, UK, Birkhäuser Boston, 2002
PDF, PS, Bibtex
2001 Tracking with a novel pose estimation algorithm
Rosenhahn B., Krüger N., Rabsch T., Sommer G.
In R. Klette, S. Peleg and G. Sommer, editors, International Workshop ``Robot Vision 2001ŽŽ Auckland, New-Zealand, Vol. 1998 of LNCS, pp. 9-18. Springer-Verlag, 2001
PDF, PS, Bibtex
2000 Pose estimation in the language of kinematics
Rosenhahn B., Zhang Y., Sommer G.
In G. Sommer and Y. Zeevi, editors, 2nd International Workshop on Algebraic Frames for the Perception-Action Cycle, AFPAC 2000, Kiel , Vol. 1888 of LNCS, pp. 284-293. Springer-Verlag, 2000
PDF, PS, Bibtex
2000 Extended Kalman filter design for motion estimation by point and line observations
Zhang Y., Rosenhahn B., Sommer G.
In G. Sommer and Y. Zeevi, editors, 2nd International Workshop on Algebraic Frames for the Perception-Action Cycle, AFPAC 2000, Kiel , Vol. 1888 of LNCS, pp. 339-348. Springer-Verlag, 2000
PDF, Bibtex