9.40. Compute Moment Invariants (2D)

Group (Subgroup)

Statistics (Statistics)

Description

This Filter computes the 2D Omega-1 and Omega 2 values from the Central Moments matrix and optionally will normalize the values to a unit circle and also optionally save the Central Moments matrix as a DataArray to the Cell Feature Attribute Matrix. Based off the paper by MacSleyne et. al [1], the algorithm will calculate the 2D central moments for each feature starting at feature id = 1. Because feature id 0 is of special significance and typically is a matrix or background it is ignored in this filter. If any feature id has a Z Delta of > 1, the feature will be skipped. This algorithm works strictly in the XY plane and is meant to be applied to a 2D image. Using the research from the cited paper certain shapes can be detected using the Omega-1 and Omega-2 values. An example usage is finding elliptical shapes in an image:

See below figure from [1] that can help the user classify objects.

Input Cell Data

Parameter Name	Parameter Type	Parameter Notes	Description
2D Image Geometry	Geometry Selection	Image	The path to the 2D image geometry to be used as input
Cell Feature Ids	Array Selection	Allowed Types: int32 Comp. Shape: 1	Data Array that specifies to which Feature each Element belongs
Feature Rect	Array Selection	Allowed Types: uint32 Comp. Shape: 6	Array holding the min xy and max xy pixel coordinates of each feature id
Normalize Moment Invariants	Bool		Should the algorithm normalize the results to unit circle.

Output Feature Data

Parameter Name	Parameter Type	Parameter Notes	Description
Feature Attribute Matrix	AttributeMatrixSelection		The path to the cell feature attribute matrix where the created data arrays will be stored
Omega 1	DataObjectName		Omega1 value
Omega 2	DataObjectName		Omega2 value
Save Central Moments	Bool		Write the Central Moments to a new Data Array
Central Moments	DataObjectName		Central Moments value

9.41. Citations

[1] J.P. MacSleyne, J.P. Simmons, M. De Graef, On the use of 2-D moment invariants for the automated classification of particle shapes, Acta Materialia, Volume 56, Issue 3, February 2008, Pages 427-437, ISSN 1359-6454, http://dx.doi.org/10.1016/j.actamat.2007.09.039. http://www.sciencedirect.com/science/article/pii/S1359645407006702

9.42. Acknowledgements

The authors would like to thank Dr. Marc De Graef from Carnegie Mellon University for enlightening discussions and a concrete implementation from which to start this filter.

Example Pipelines

License & Copyright

Please see the description file distributed with this plugin.

DREAM3D-NX Help

If you need help, need to file a bug report or want to request a new feature, please head over to the DREAM3DNX-Issues GitHub site where the community of DREAM3D-NX users can help answer your questions.