Matching Planar Features for Robot Localization

Localizing a vehicle with a vision based system often requires to match and track landmarks whose position is known. This paper tries to define a new method to track some features in modeling them as local planar patches with a monocular camera. First a learning sequence is recorded to compute the planar features and their orientation around landmarks tracked on several views. Then in the localization part, camera pose is predicted and features are transformed to fit with the scene as seen by the camera. Landmarks can then easily be matched and position is computed more accurately. With this method many features can be tracked on longer sequences than with standard methods, even if the camera is moving away from the learning trajectory. This improves the localization.

[1]  Helder Araújo,et al.  A Fully Projective Formulation to Improve the Accuracy of Lowe's Pose-Estimation Algorithm , 1998, Comput. Vis. Image Underst..

[2]  Vincent Lepetit,et al.  Keypoint Signatures for Fast Learning and Recognition , 2008, ECCV.

[3]  Luc Van Gool,et al.  SURF: Speeded Up Robust Features , 2006, ECCV.

[4]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[5]  Andrew J. Davison,et al.  Active Matching , 2008, ECCV.

[6]  Andrew Zisserman,et al.  Multiple View Geometry in Computer Vision (2nd ed) , 2003 .

[7]  Andrew W. Fitzgibbon,et al.  Bundle Adjustment - A Modern Synthesis , 1999, Workshop on Vision Algorithms.

[8]  Antti Oulasvirta,et al.  Computer Vision – ECCV 2006 , 2006, Lecture Notes in Computer Science.

[9]  Michel Dhome,et al.  Monocular Vision for Mobile Robot Localization and Autonomous Navigation , 2007, International Journal of Computer Vision.

[10]  Wolfram Burgard,et al.  Robust Monte Carlo localization for mobile robots , 2001, Artif. Intell..

[11]  Ian D. Reid,et al.  Locally Planar Patch Features for Real-Time Structure from Motion , 2004, BMVC.

[12]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[13]  Vincent Lepetit,et al.  Randomized trees for real-time keypoint recognition , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[14]  Simon Lacroix,et al.  Using planar facets for stereovision SLAM , 2008, 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems.