Authors:
Hildegard Kuehne
1
and
Annika Woerner
2
Affiliations:
1
University of Karlsruhe(TH), Germany
;
2
Institut for Algorithms and Cognitive Systems, University of Karlsruhe (TH), Germany
Keyword(s):
Motion segmentation, Articulated body tracking, Motion recognition.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Computer Vision, Visualization and Computer Graphics
;
Human-Computer Interaction
;
Image and Video Analysis
;
Methodologies and Methods
;
Motion and Tracking
;
Motion, Tracking and Stereo Vision
;
Pattern Recognition
;
Physiological Computing Systems
;
Segment Cluster Tracking
;
Segmentation and Grouping
Abstract:
The correct segmentation of articulated motion is an important factor to extract and understand the functional structures of complex, articulated objects. Segmenting such body motion without additional appearance information is still a challenging task, because articulated objects as e.g. the human body are mainly based on fine, connected structures. The proposed approach combines consensus based motion segmentation with biological inspired visual perception criteria. This allows the grouping of sparse, dependent moving features points into several clusters, representing the rigid elements of an articulated structure. It is shown how geometric and time-based feature properties can be used to improve the result of motion segmentation in this context. We evaluated our algorithm on artificial as well as natural video sequences in order to segment the motion of human body elements. The results of the evaluation of parameter influences and also the practical evaluation show, that good mot
ion segmentation can be achieved by this approach.
(More)