Object contour extraction based on charged snake model

Liu, Man (2011) Object contour extraction based on charged snake model. Masters thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (10Mb)

Abstract

The active contour model, or snake, is one of the most successful variational models in image extraction and segmentation. In this thesis, a novel Charged Snake Model (CSM), based on electrostatic theory for object contour extraction is proposed. This method overcomes several difficulties existing in conventional parametric snake models. A closed contour locates the initial contour of the snake and each point in the initial snake contour is regarded as a charge, which makes the initial CSM snake close enough to the object boundary to allow for faster convergence. Furthermore, due to the interaction among all charges in the snake, the snake model is not sensitive to and is not influenced by the initialization position. As for CSM snake, an improved and associated energy function, with employing additional parameters, is generated. Under the influence of internal and external image dependent forces, the initial CSM snake deforms towards the minimum of the energy function where the object boundary is located and the CSM snake reaches its convergence. By this process, a complete object shape, as well as the object position described by the CSM snake, can be obtained. This shape and position information can then be used in further shape analysis.

Item Type: Thesis (Masters)
URI: http://research.library.mun.ca/id/eprint/9569
Item ID: 9569
Additional Information: Bibliography: leaves 66-71.
Department(s): Science, Faculty of > Computer Science
Date: 2011
Date Type: Submission
Library of Congress Subject Heading: Image processing--Digital techniques; Pattern perception--Computer simulation; Pattern recognition systems

Actions (login required)

View Item View Item

Downloads

Downloads per month over the past year

View more statistics