Output details
11 - Computer Science and Informatics
Kingston University
Return to search
Output 0 of 0 in the submission
Output title
Tracking People across Multiple Non-overlapping RGB-D Sensors
Type
E - Conference contribution
Name of conference/published proceedings
2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
Volume number
-
Issue number
-
First page of article
831
ISSN of proceedings
2160-7508
Year of publication
2013
Number of additional authors
1
Additional information
<23> While the Microsoft Kinect depth sensing device has triggered an enormous amount of new research in computer vision, this paper represents the first contribution in combining the depth data from multiple Kinect sensors. A method of establishing the inter-sensor homographies of these necessarily non-overlapping fields of view was developed. Within a surveillance context, the object detection and tracking is now computed on the ground plane which we show has a dramatic impact on the resolution of static and dynamic occlusions.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Citation count
0
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
