Documentation/Nightly/Modules/SobolevSegmenter

Introduction and Acknowledgements

This work is part of the National Alliance for Medical Image Computing (NA-MIC), funded by the National Institutes of Health through the NIH Roadmap for Medical Research. Information on NA-MIC can be obtained from the NA-MIC website. Author: Arie Nakhmani, UAB Contributor1: Allen Tannenbaum, UAB Contact: Arie Nakhmani, <email>nakhmani@gmail.com</email>

University of Alabama at Birmingham

Module Description

This extension implements Sobolev inner product based active contour, using Chan-Vese energy functional. The segmentation is appropriate for 2D images. The obtained parametric contour is generally smooth, but able to catch concavities.

Use Cases

The Sobolev segmenter is a general image segmenter, and it can be used with any 2D data, as explained in the tutorial.

Tutorials

1. Load the image (input volume):
2. Use built in editor to select a single initial mask (or load a binary mask file):
3. Select Segmenation->SobolevSegmenter module
4. Choose the Input Volume and the Initial Mask accordingly. Create a new volume for the Output Volume.
5. Press Apply button.
6. After a few second the following output volume should appear:

Panels and their use

The module has the following panel:

The IO section of this panel defines two input images (data and initial mask) and one output image (final mask). The algorithm has three parameters: self-explanatory number of iterations and contour evolution step size. In addition, the parameter lambda chooses the smoothness of the contour (smoothing kernel width).

Similar Modules

N/A

References

N/A