Adaptive patch-based background modelling for improved foreground object segmentation and tracking

Reddy, Vikas, Sanderson, Conrad, Sanin, Andres and Lovell, Brian (2010). Adaptive patch-based background modelling for improved foreground object segmentation and tracking. In: Proceedings. IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS 2010). 2010 7th International Conference on Advanced Video and Signal-Based Surveillance, Boston, MA, United States, (172-179). 29 August - 1 September 2010. doi:10.1109/AVSS.2010.84

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
reddy_adaptive_avss_2010.pdf Full-text from publisher Click to show the corresponding preview/stream application/pdf 620.19KB 4

Author Reddy, Vikas
Sanderson, Conrad
Sanin, Andres
Lovell, Brian
Title of paper Adaptive patch-based background modelling for improved foreground object segmentation and tracking
Conference name 2010 7th International Conference on Advanced Video and Signal-Based Surveillance
Conference location Boston, MA, United States
Conference dates 29 August - 1 September 2010
Proceedings title Proceedings. IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS 2010)
Place of Publication Piscataway, NJ, United States
Publisher IEEE Computer Society
Publication Year 2010
Sub-type Fully published paper
DOI 10.1109/AVSS.2010.84
ISBN 9781424483105
Start page 172
End page 179
Total pages 8
Collection year 2011
Language eng
Abstract/Summary A robust foreground object segmentation technique is proposed, capable of dealing with image sequences containing noise, illumination variations and dynamic backgrounds. The method employs contextual spatial information by analysing each image on an overlapping patch-by-patch basis and obtaining a low-dimensional texture descriptor for each patch. Each descriptor is passed through an adaptive multi-stage classifier, comprised of a likelihood evaluation, an illumination robust measure, and a temporal correlation check. A probabilistic foreground mask generation approach integrates the classification decisions by exploiting the overlapping of patches, ensuring smooth contours of the foreground objects as well as effectively minimising the number of errors. The parameter settings are robust against wide variety of sequences and post-processing of foreground masks is not required. Experiments on the difficult Wallflower and I2R datasets show that the proposed method obtains considerably better results (both qualitatively and quantitatively) than methods based on Gaussian mixture models, feature histograms, and normalised vector distances. Further experiments on the CAVIAR dataset (using several tracking algorithms) indicate that the proposed method leads to considerable improvements in object tracking accuracy. © 2010 IEEE.
Keyword Gaussian processes
Image classification
Image segmentation
Image sequences
Image texture
Probability
Q-Index Code E1
Q-Index Status Confirmed Code
Institutional Status UQ
Additional Notes Session Foreground/Background Segmentation I

 
Versions
Version Filter Type
Citation counts: Scopus Citation Count Cited 10 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Access Statistics: 146 Abstract Views, 7 File Downloads  -  Detailed Statistics
Created: Fri, 26 Nov 2010, 18:17:06 EST by Conrad Sanderson on behalf of School of Information Technol and Elec Engineering