Grant: Discovery Grants Program - Individual
University of Alberta
$25,000.00
Feature correspondence for image analysis
EF 2010-11
Résumé
Feature Correspondence for Image Analysis
Feature correspondence (FC) is one of the most difficult core problems in image analysis and computer vision. Given two sets of features, FC aims at finding one-to-one matching or correspondences between them. Applications of FC are numerous: image registration, robot navigation, optical flow, stereo depth estimation, visual object tracking, and many more. State-of-the-art solutions to FC fail due mainly to three factors: (a) large-scale problem (i.e., number of features is very large), (b) excessive outliers (many features with no match), and (c) severe measurement noise. These three factors are often present in significant amounts in the aforementioned applications. My proposed solution to FC is based on optimization of a concave cost function taking adequate care of all of these three factors. Concave function minimization has a rich history in operations research and has satisfactorily solved many applications ranging from economics, control theory, structural engineering, production process, and so on. Along with the aforementioned large-scale applications, the proposed research will be able to solve some novel nano image visualization problems.
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
341552-2010
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
2010
Competition year année de concours
2010
Award amount
25000
Program id
RGPIN
Program name 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
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
1206
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Engineering
Area of application fr
Génie
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
1
Nb partie
5
Application title
Feature correspondence for image analysis
Keyword
NSERC
Application summary
Feature Correspondence for Image Analysis
Feature correspondence (FC) is one of the most difficult core problems in image analysis and computer vision. Given two sets of features, FC aims at finding one-to-one matching or correspondences between them. Applications of FC are numerous: image registration, robot navigation, optical flow, stereo depth estimation, visual object tracking, and many more. State-of-the-art solutions to FC fail due mainly to three factors: (a) large-scale problem (i.e., number of features is very large), (b) excessive outliers (many features with no match), and (c) severe measurement noise. These three factors are often present in significant amounts in the aforementioned applications. My proposed solution to FC is based on optimization of a concave cost function taking adequate care of all of these three factors. Concave function minimization has a rich history in operations research and has satisfactorily solved many applications ranging from economics, control theory, structural engineering, production process, and so on. Along with the aforementioned large-scale applications, the proposed research will be able to solve some novel nano image visualization problems.