p.986
p.990
p.994
p.1000
p.1004
p.1011
p.1018
p.1022
p.1027
A Novel Approach for Active Contour Initialization in 3D Medical Image Segmentation
Abstract:
Contour initialization is a big problem of the active contour model. Using the continuous features of the three-dimensional medical image, this paper proposes an initial contour prediction model. There are some changes in the boundary contours of the same object. We attribute these changes to continuous translation and similar deformation, and quantify into the centroid displacement and displacement of the point in the direction of Normal. The curve fitting method is used to predict the centroid displacement and the displacement of the points of the contours, which can provide more accurate prediction of changes in the contour. By predicting the initial contour, we have solved the contour initialization problem of the parametric active contour with external force using vector field convolution.
Info:
Periodical:
Pages:
1004-1010
Citation:
Online since:
December 2012
Authors:
Price:
Сopyright:
© 2013 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: