Improved Rodent Contour Extraction Using A Priori Shape Information

Rowland R. Sillito; T.C. Lukins; J. D. Armstrong

Improved Rodent Contour Extraction Using A Priori Shape Information

Rowland R. Sillito, T.C. Lukins, J. D. Armstrong

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We propose a method for automatically setting the foreground detection threshold implicit in background subtraction algorithms by measuring the similarity between the shape of a detected foreground region and a set of reference contours over a range of thresholds, and selecting the threshold that maximises this similarity measure. This method is shown to select appropriate thresholds for a range of unseen video frames.

Original language	English
Title of host publication	Workshop on the Visual Observation and Analysis of Animal and Insect Behavior (held at ICPR2010)
Number of pages	4
Publication status	Published - 2010

Access to Document

http://homepages.inf.ed.ac.uk/rbf/VAIB10PAPERS/rsvaib_paper.pdf

Cite this

@inproceedings{c0e4c19e61434085baedc7cbf617b1a6,

title = "Improved Rodent Contour Extraction Using A Priori Shape Information",

abstract = "We propose a method for automatically setting the foreground detection threshold implicit in background subtraction algorithms by measuring the similarity between the shape of a detected foreground region and a set of reference contours over a range of thresholds, and selecting the threshold that maximises this similarity measure. This method is shown to select appropriate thresholds for a range of unseen video frames.",

author = "Sillito, {Rowland R.} and T.C. Lukins and Armstrong, {J. D.}",

year = "2010",

language = "English",

booktitle = "Workshop on the Visual Observation and Analysis of Animal and Insect Behavior (held at ICPR2010)",

}

TY - GEN

T1 - Improved Rodent Contour Extraction Using A Priori Shape Information

AU - Sillito, Rowland R.

AU - Lukins, T.C.

AU - Armstrong, J. D.

PY - 2010

Y1 - 2010

N2 - We propose a method for automatically setting the foreground detection threshold implicit in background subtraction algorithms by measuring the similarity between the shape of a detected foreground region and a set of reference contours over a range of thresholds, and selecting the threshold that maximises this similarity measure. This method is shown to select appropriate thresholds for a range of unseen video frames.

AB - We propose a method for automatically setting the foreground detection threshold implicit in background subtraction algorithms by measuring the similarity between the shape of a detected foreground region and a set of reference contours over a range of thresholds, and selecting the threshold that maximises this similarity measure. This method is shown to select appropriate thresholds for a range of unseen video frames.

M3 - Conference contribution

BT - Workshop on the Visual Observation and Analysis of Animal and Insect Behavior (held at ICPR2010)

ER -