Deakin home > Deakin University Library > Deakin Research Online > Vision based target tracking using robust linear filtering

Vision based target tracking using robust linear filtering

Bishop, Adrian, Savkin, Andrey and Pathirana, Pubudu 2007, Vision based target tracking using robust linear filtering, in ECC '07: proceedings of the European Control Conference, European Union Control Association (EUCA), [Kos, Greece], pp. 1442-1447.

Attached Files (Some files may be inaccessible until you login with your Deakin Research Online credentials)
Name Description MIMEType Size Downloads

Title Vision based target tracking using robust linear filtering
Author(s) Bishop, Adrian
Savkin, Andrey
Pathirana, Pubudu
Conference name European Control Conference ( 2007: Kos, Greece)
Conference location Kos, Greece
Conference dates 2-5 July 2007
Title of proceedings ECC '07: proceedings of the European Control Conference
Editor(s) [Unknown]
Publication date 2007
Conference series European Control Conference
Start page 1442
End page 1447
Publisher European Union Control Association (EUCA)
Place of publication [Kos, Greece]
Keyword(s) robust control
estimation
Kalman filtering
Summary The use of perspective projection in tracking a target from a video stream involves nonlinear observations. The target dynamics, however, are modeled in Cartesian coordinates and result in a linear system. In this paper we provide a robust version of a linear Kalman filter and perform a measurement conversion technique on the nonlinear optical measurements. We show that our linear robust filter significantly outperforms the Extended Kalman Filter. Moreover, we prove that the state estimation error is bounded in a probabilistic sense.
ISBN 9789608902855
9608902851
Language eng
Field of Research 080104 Computer Vision
HERDC Research category E1 Full written paper - refereed
ERA Research output type E Conference publication
Persistent URL http://hdl.handle.net/10536/DRO/DU:30008057

Document type: Conference Paper
Collection: School of Engineering and Information Technology
Connect to link resolver
 
Unless expressly stated otherwise, the copyright for items in Deakin Research Online is owned by the author, with all rights reserved.

Versions
Version Filter Type
Access Statistics: 410 Abstract Views, 0 File Downloads  -  Detailed Statistics
Created: Mon, 29 Sep 2008, 09:04:02 EST