Share Email Print
cover

Proceedings Paper

Tracking moving objects across non-overlapping cameras
Author(s): Isaac Cohen; Yunqian Ma; Ben Miller
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In this paper we present an approach for tracking people across non overlapping cameras. The approach proposed is based a multi-dimensional feature vector and its covariance, defining an appearance model of every detected blob in the network of cameras. The model integrates relative position, color and texture descriptors of each detected object. Association of objects across non-overlapping cameras is performed by matching detected objects appearance with past observations. Availability of tracking within every camera can further improve the accuracy of such association by matching several targets appearance models with detected regions. For this purpose we present an automatic clustering technique allowing to build a multi-valued appearance model from a collection of covariance matrices. The proposed approach does not require geometric or colorimetric calibration of the cameras. We will illustrate the method for tracking people in relatively crowded scenes in a collection of indoors cameras taken in a mass transportation site. We will present success and challenges yet to be addressed by the proposed approach.

Paper Details

Date Published: 8 October 2007
PDF: 8 pages
Proc. SPIE 6741, Optics and Photonics for Counterterrorism and Crime Fighting III, 674107 (8 October 2007); doi: 10.1117/12.737648
Show Author Affiliations
Isaac Cohen, Honeywell Technology (United States)
Yunqian Ma, Honeywell Technology (United States)
Ben Miller, Honeywell Technology (United States)


Published in SPIE Proceedings Vol. 6741:
Optics and Photonics for Counterterrorism and Crime Fighting III
Colin Lewis, Editor(s)

© SPIE. Terms of Use
Back to Top