Integration of Model-based and Model-free Cues for Visual Object Tracking in 3D
2005 (English)In: 2005 IEEE International Conference on Robotics and Automation (ICRA), IEEE Computer Society, 2005, 1554-1560 p.Conference paper (Refereed)
Vision is one of the most powerful sensory modalities in robotics, allowing operation in dynamic environments. One of our long-term research interests is mobile manipulation, where precise location of the target object is commonly required during task execution. Recently, a number of approaches have been proposed for real-time 3D tracking and most of them utilize an edge (wireframe) model of the target. However, the use of an edge model has significant problems in complex scenes due to occlusions and multiple responses, especially in terms of initialization. In this paper, we propose a new tracking method based on integration of model-based cues with automatically generated model-free cues, in order to improve tracking accuracy and to avoid weaknesses of edge based tracking. The integration is performed in a Kalman filter framework that operates in real-time. Experimental evaluation shows that the inclusion of model-free cues offers superior performance.
Place, publisher, year, edition, pages
IEEE Computer Society, 2005. 1554-1560 p.
, IEEE International Conference on Robotics and Automation, ISSN 2152-4092 ; 2005
model-based tracking, model-free track ing, cue integration, iterated extended Kalman filter
Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-63813DOI: 10.1109/ROBOT.2005.1570335ISI: 000235460101057ScopusID: 2-s2.0-33846133573ISBN: 0-7803-8914-XOAI: oai:DiVA.org:kth-63813DiVA: diva2:482704
2005 IEEE International Conference on Robotics and Automation; Barcelona; 18 April 2005 through 22 April 2005
QC 201202032012-01-242012-01-242012-02-03Bibliographically approved