itk::AdvancedKappaStatisticImageToImageMetric< TFixedImage, TMovingImage > Class Template Reference

#include <itkAdvancedKappaStatisticImageToImageMetric.h>

Detailed Description

template<class TFixedImage, class TMovingImage>
class itk::AdvancedKappaStatisticImageToImageMetric< TFixedImage, TMovingImage >

Computes similarity between two objects to be registered.

This class is templated over the type of the fixed and moving images to be compared. The metric here is designed for matching pixels in two images with the same exact value. Only one value can be considered (the default is 255) and can be specified with the SetForegroundValue method. In the computation of the metric, only foreground pixels are considered. The metric value is given by 2*|A&B|/(|A|+|B|), where A is the foreground region in the moving image, B is the foreground region in the fixed image, & is intersection, and |.| indicates the area of the enclosed set. The metric is described in "Morphometric Analysis of White Matter Lesions in MR Images: Method and Validation", A. P. Zijdenbos, B. M. Dawant, R. A. Margolin, A. C. Palmer.

This metric is especially useful when considering the similarity between binary images. Given the nature of binary images, a nearest neighbor interpolator is the preferred interpolator.

Metric values range from 0.0 (no foreground alignment) to 1.0 (perfect foreground alignment). When dealing with optimizers that can only minimize a metric, use the ComplementOn() method.

Definition at line 56 of file itkAdvancedKappaStatisticImageToImageMetric.h.

Inheritance diagram for itk::AdvancedKappaStatisticImageToImageMetric< TFixedImage, TMovingImage >:

Data Structures
struct	KappaGetValueAndDerivativePerThreadStruct
struct	MultiThreaderAccumulateDerivativeType

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::CoordinateRepresentationType	CoordinateRepresentationType