On the merits of the Gaussian Mixture as a model for oriented edgel distributions

Assheton, Philip and Hunter, Andrew (2008) On the merits of the Gaussian Mixture as a model for oriented edgel distributions. Technical Report. University of Lincoln. (Unpublished)

Documents
On the Merits of the Gaussian Mixture as a Model for Oriented Edgel Distributions
This paper provides a theoretical and empirical justification for the use of Gaussian Mixture Models in modelling oriented edgel distributions of rigid and natural objects.
[img]
[Download]
[img]
Preview
PDF
GMMEdgelDistributionsAsshetonHunter.pdf

496kB

Abstract

The aim of this report is to establish the credibility of the Gaussian Mixture Model (GMM) as a model for the distributions of oriented edgels of rigid and biological objects in noisy images. This is tackled in two stages: first, the response of the Soble filter to noisy pixels is analysed to show that the result holds for smooth ridid objects. Second, arguments are presented to support the proposition that the model can also effectively capture the added uncertainty introduced by natural shape variation, as found in images of biological objects. The result has particular application in the extension of the Generalized Hough Transform (GHT) to deformable shapes; in particular if offers a tailored and manipulable alternative to the non-parametric kernel density estimate used by Ecabert and Thiran.

Item Type:Paper or Report (Technical Report)
Additional Information:The aim of this report is to establish the credibility of the Gaussian Mixture Model (GMM) as a model for the distributions of oriented edgels of rigid and biological objects in noisy images. This is tackled in two stages: first, the response of the Soble filter to noisy pixels is analysed to show that the result holds for smooth ridid objects. Second, arguments are presented to support the proposition that the model can also effectively capture the added uncertainty introduced by natural shape variation, as found in images of biological objects. The result has particular application in the extension of the Generalized Hough Transform (GHT) to deformable shapes; in particular if offers a tailored and manipulable alternative to the non-parametric kernel density estimate used by Ecabert and Thiran.
Keywords:Hough transform, shape-based object detection, Mixture of Gaussians
Subjects:G Mathematical and Computer Sciences > G740 Computer Vision
Divisions:College of Science > School of Computer Science
ID Code:1501
Deposited By: Andrew Hunter
Deposited On:14 May 2008 07:36
Last Modified:13 Mar 2013 08:27

Repository Staff Only: item control page