Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

University of Alberta

$20,000.00

Hybrid computational strategies for image segmentation and object tracking

EF 2009-10

Résumé
Image segmentation and object tracking are two fundamental tasks upon which several automated image analysis and computer vision applications rely. Ideally, image segmentation and object tracking algorithms should make use of available information as much as possible. The information about the objects of interest or segments typically can be geometric or topological (e.g., shape, size, solidity or hollowness etc), as well as some image-based characteristics (e.g., intensity/color/texture). Spatial point process techniques, a tool-set from probability and statistics, utilize these pieces of information in a theoretically sound Bayesian framework. However, the drawback of point process techniques is non-efficient computation. On the deterministic side, level set methods, which were originally introduced as numerical techniques for front propagation, are computationally efficient in image segmentation and object tracking. However, level set methods are not as versatile as the point process methods in terms of utilizing the information about the objects/segments. In this research I aim to merge these two fascinating image analysis tools together to complement each other. This marriage of the two methods is a novel concept and is proposed in a theoretically sound stochastic sampling framework. The anticipated outcome of this research is a computationally efficient yet generic image segmentation and object tracking tool.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Alberta, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
341552-2007
Name nom
Ray, Nilanjan
Department département
Computing Science
Organization id
9
Institution établissement
University of Alberta
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2009
Competition year année de concours
2007
Award amount
20000
Program id
RGPIN
Program nane en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
331
Committee name en
Computing and Information Sciences - B
Committee name fr
Informatique - B
Area of application code
301
Area of application group en
Energy resources
Area of application group fr
Ressources d'énergie
Area of application en
Oil, gas and coal
Area of application fr
Pétrole, gaz et charbon
Research subject code
2708
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Image processing
Research subject fr
Traitement des images
Partie
3
Nb partie
3
Application title
Hybrid computational strategies for image segmentation and object tracking
Keyword
NA
Application summary
Image segmentation and object tracking are two fundamental tasks upon which several automated image analysis and computer vision applications rely. Ideally, image segmentation and object tracking algorithms should make use of available information as much as possible. The information about the objects of interest or segments typically can be geometric or topological (e.g., shape, size, solidity or hollowness etc), as well as some image-based characteristics (e.g., intensity/color/texture). Spatial point process techniques, a tool-set from probability and statistics, utilize these pieces of information in a theoretically sound Bayesian framework. However, the drawback of point process techniques is non-efficient computation. On the deterministic side, level set methods, which were originally introduced as numerical techniques for front propagation, are computationally efficient in image segmentation and object tracking. However, level set methods are not as versatile as the point process methods in terms of utilizing the information about the objects/segments. In this research I aim to merge these two fascinating image analysis tools together to complement each other. This marriage of the two methods is a novel concept and is proposed in a theoretically sound stochastic sampling framework. The anticipated outcome of this research is a computationally efficient yet generic image segmentation and object tracking tool.