Chan & Vese

Class: NodeChanVeseSegmentation

Node Icon

The Chan-Vese segmentation algorithm is designed to segment objects without clearly defined boundaries. This algorithm is based on level sets that are evolved iteratively to minimize an energy, which is defined by weighted values corresponding to the sum of differences intensity from the average value outside the segmented region, the sum of differences from the average value inside the segmented region, and a term which is dependent on the length of the boundary of the segmented region.

Typical values for \(\lambda_1\) and \(\lambda_2\) are 1. If the ‘background’ is very different from the segmented object in terms of distribution (for example, a uniform black image with figures of varying intensity), then these values should be different from each other.

This algorithm was first proposed by Tony Chan and Luminita Vese, in a publication entitled “An Active Contour Model Without Edges”, see 1.

Example Workflows

Chan Vese segmentation example

Inputs

Initial Level Set

The initial level set from which to start.

Type: Image4DFloat, Required, Single

Image

The image to be segmented.

Type: Image4DFloat, Required, Single

Outputs

Output

The segmented image.

Type: Image4DFloat

Settings

Area Weight Number

Area regularization values.

Curvature Weight Number

Scales all curvature weight values.

Epsilon Number

Width of regularization of Heaviside function.

Lambda 1 Number

Internal intensity difference weight.

Lambda 2 Number

External intensity difference weight.

Reinitialization Smoothing Weight Number

Weight of the laplacian smoothing term.

Volume Number

Volume.

Volume Matching Weight Number

Volume matching weight.

Use Image Spacing Boolean

Set whether or not the filter will use the spacing of the input image in its calculations.

Maximum RMS Error Number

Value of RMS change below which the filter should stop. This is a convergence criterion.

Iterations Integer

Set the number of iterations.

References

“An active contour model without edges” T.Chan and L.Vese.In Scale - Space Theories in Computer Vision, pages 141 - 151, 1999.
“Cell Tracking using Coupled Active Surfaces for Nuclei and Membranes” http://www.insight-journal.org/browse/publication/642 https://hdl.handle.net/10380/3055
Chan Vese Segmentation in SimpleITK