![](/images/graphics-bg.png)
Mixed Signature: An Invariant Descriptor for 3D Motion Trajectory Perception and Recognition
Joint Authors
Li, Y. F.
Yang, Jianyu
Wang, Keyi
Wu, Yuan
Altieri, Giuseppe
Scalia, Massimo
Source
Mathematical Problems in Engineering
Issue
Vol. 2012, Issue 2012 (31 Dec. 2012), pp.1-29, 29 p.
Publisher
Hindawi Publishing Corporation
Publication Date
2010-08-01
Country of Publication
Egypt
No. of Pages
29
Main Subjects
Abstract EN
Motion trajectory contains plentiful motion information of moving objects, for example, human gestures and robot actions.
Motion perception and recognition via trajectory areuseful for characterizing them and a flexible descriptor of motion trajectory plays importantrole in motion analysis.
However, in the existing tasks, trajectories were mostly used in rawdata and effective descriptor is lacking.
In this paper, we present a mixed invariant signaturedescriptor with global invariants for motion perception and recognition.
The mixed signatureis viewpoint invariant for local and global features.
A reliable approximation of the mixedsignature is proposed to reduce the noise in high-order derivatives.
We use this descriptor formotion trajectory description and explore the motion perception with DTW algorithm forsalient motion features.
To achieve better accuracy, we modified the CDTW algorithm fortrajectory matching in motion recognition.
Furthermore, a controllable weight parameter isintroduced to adjust the global features for tasks in different circumstances.
The conductedexperiments validated the proposed method.
American Psychological Association (APA)
Yang, Jianyu& Li, Y. F.& Wang, Keyi& Wu, Yuan& Altieri, Giuseppe& Scalia, Massimo. 2010. Mixed Signature: An Invariant Descriptor for 3D Motion Trajectory Perception and Recognition. Mathematical Problems in Engineering،Vol. 2012, no. 2012, pp.1-29.
https://search.emarefa.net/detail/BIM-1029644
Modern Language Association (MLA)
Yang, Jianyu…[et al.]. Mixed Signature: An Invariant Descriptor for 3D Motion Trajectory Perception and Recognition. Mathematical Problems in Engineering No. 2012 (2012), pp.1-29.
https://search.emarefa.net/detail/BIM-1029644
American Medical Association (AMA)
Yang, Jianyu& Li, Y. F.& Wang, Keyi& Wu, Yuan& Altieri, Giuseppe& Scalia, Massimo. Mixed Signature: An Invariant Descriptor for 3D Motion Trajectory Perception and Recognition. Mathematical Problems in Engineering. 2010. Vol. 2012, no. 2012, pp.1-29.
https://search.emarefa.net/detail/BIM-1029644
Data Type
Journal Articles
Language
English
Notes
Includes bibliographical references
Record ID
BIM-1029644