Localization with omnidirectional images using the radial trifocal tensor

In this paper we present a technique to linearly recover 2D structure and motion in man made environments from three uncalibrated omnidirectional views. We use vertical lines from the scene which are projected as radial lines in the images and are automatically matched. The algorithm is based on a 1D radial trifocal tensor which encodes the relations of the three views and the projected lines. We include experiments with real images, which demonstrate the good performance of the method and its application to robotic tasks, such as robot localization based in a database of reference images or to obtain the initial values of robot and landmarks localization in SLAM algorithms

[1]  Olivier D. Faugeras,et al.  Self-Calibration of a 1D Projective Camera and Its Application to the Self-Calibration of a 2D Projective Camera , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Marc Pollefeys,et al.  The radial trifocal tensor: a tool for calibrating the radial distortion of wide-angle cameras , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[3]  Peter I. Corke,et al.  Omnidirectional visual odometry for a planetary rover , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[4]  Emanuele Menegatti,et al.  Testing omnidirectional vision-based Monte Carlo localization under occlusion , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[5]  Luc Van Gool,et al.  Hierarchical localization by matching vertical lines in omnidirectional images , 2006 .

[6]  Frank Dellaert,et al.  Linear 2D localization and mapping for single and multiple robot scenarios , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[7]  Magnus Oskarsson,et al.  Solutions and Ambiguities of the Structure and Motion Problem for 1D Retinal Vision , 2000, Journal of Mathematical Imaging and Vision.

[8]  Peter J. Rousseeuw,et al.  Robust Regression and Outlier Detection , 2005, Wiley Series in Probability and Statistics.

[9]  Yasushi Yagi,et al.  Iconic memory-based omnidirectional route panorama navigation , 2003, 2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422).

[10]  Shree K. Nayar,et al.  Ego-motion and omnidirectional cameras , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[11]  Michael Werman,et al.  Trilinearity of three perspective views and its associated tensor , 1995, Proceedings of IEEE International Conference on Computer Vision.

[12]  Dana Cobzas,et al.  D Robot Localization with Image-Based Panoramic Models Using Vertical Line Features , 2000 .

[13]  Kostas Daniilidis,et al.  Structure and motion from uncalibrated catadioptric views , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[14]  Josechu J. Guerrero,et al.  Robust Line Matching and Estimate of Homographies Simultaneously , 2003, IbPRIA.