Spatio-temporal modeling in the farmyard domain

D. R. Magee, R. D. Boyle

Research output: Chapter in Book/Report/Conference proceedingConference Proceeding (Non-Journal item)

2 Citations (SciVal)

Abstract

A temporal modelling and prediction scheme based on modelling a ‘history space’ using Gaussian mixture models is presented. A point in this space represents an abstraction of a complete object history as opposed to finite histories used in Markov methods. It is shown how this ‘History Space Classifier’ may be incorporated into an existing scheme for spatial object modelling and tracking to improve tracking speed and robustness and to classify object ‘behaviour’ into normal and abnormal. An application to the tracking and monitoring of livestock is also presented in this paper.

Original languageEnglish
Title of host publicationArticulated Motion and Deformable Objects - 1st International Workshop, AMDO 2000, Proceedings
EditorsHans-Hellmut Nagel, Francisco J. Perales Lopez
PublisherSpringer Nature
Pages83-95
Number of pages13
ISBN (Print)354067912X
DOIs
Publication statusPublished - 2000
Externally publishedYes
Event1st International Workshop on Articulated Motion and Deformable Objects, AMDO 2000 - Palma de Mallorca, Spain
Duration: 07 Sept 200009 Sept 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1899
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st International Workshop on Articulated Motion and Deformable Objects, AMDO 2000
Country/TerritorySpain
CityPalma de Mallorca
Period07 Sept 200009 Sept 2000

Keywords

  • Behaviour model
  • History space classifier
  • Livestock

Fingerprint

Dive into the research topics of 'Spatio-temporal modeling in the farmyard domain'. Together they form a unique fingerprint.

Cite this