2009 | OriginalPaper | Chapter
Tracking Objects beyond Rigid Motion
Authors : Nicole Artner, Adrian Ion, Walter G. Kropatsch
Published in: Graph-Based Representations in Pattern Recognition
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Tracking multiple features of a rigid or an articulated object, without considering the underlying structure, becomes ambiguous if the target model (for example color histograms) is similar to other nearby regions or to the background. Instead of tracking multiple features independently, we propose an approach that integrates the underlying structure into the tracking process using an attributed graph. The independent tracking processes are driven to a solution that satisfies the visual as well as the structural constraints. An approach for rigid objects is presented and extended to handle articulated objects consisting of rigid parts. Experimental results on real and synthetic videos show promising results in scenes with considerable amount of occlusion.