ENH: Add Python wrapping for ObjectToObjectMultiMetricv4 with explicit TVirtualImage

[Copilot]

ObjectToObjectMultiMetricv4 was not wrapped for Python, limiting multi-metric registration to third-party libraries like ANTsPy. Additionally, ImageToImageMetricv4 classes were wrapped without explicit TVirtualImage parameter specification.

Description

This PR adds Python wrapping for the ObjectToObjectMultiMetricv4 class and updates all ImageToImageMetricv4 derived classes to include explicit TVirtualImage parameter specification, enabling multi-metric registration workflows directly in ITK Python with consistent template parameter usage.

Changes

• Added itkObjectToObjectMultiMetricv4.wrap in Modules/Registration/Metricsv4/wrapping/
• Updated all ImageToImageMetricv4 derived classes to include explicit TVirtualImage parameter:
  • itkImageToImageMetricv4.wrap
  • itkMeanSquaresImageToImageMetricv4.wrap
  • itkCorrelationImageToImageMetricv4.wrap
  • itkDemonsImageToImageMetricv4.wrap
  • itkANTSNeighborhoodCorrelationImageToImageMetricv4.wrap
  • itkJointHistogramMutualInformationImageToImageMetricv4.wrap
  • itkMattesMutualInformationImageToImageMetricv4.wrap
• Wraps dimensions 2 and 3 with both float and double types
• Explicitly specifies TVirtualImage parameter based on wrapped image types (itk::Image<float/double, dimension>)
• Added Python test itkObjectToObjectMultiMetricv4Test.py demonstrating usage:
  • Creating multi-metric instances with explicit TVirtualImage
  • Adding component metrics with explicit TVirtualImage specification
  • Setting and retrieving metric weights
  • Verifying the number of metrics

Usage Example

import itk

# Create multi-metric for combined registration
Dimension = 2
ImageType = itk.Image[itk.F, Dimension]

multi_metric = itk.ObjectToObjectMultiMetricv4[Dimension, Dimension, ImageType].New()

# Add component metrics with explicit TVirtualImage
ms_metric = itk.MeanSquaresImageToImageMetricv4[ImageType, ImageType, ImageType].New()
multi_metric.AddMetric(ms_metric)

# Set weights
weights = itk.Array[itk.D](1)
weights[0] = 0.5
multi_metric.SetMetricWeights(weights)

PR Checklist

• No API changes were made (or the changes have been approved)
• No major design changes were made (or the changes have been approved)
• Added test (or behavior not changed)
• Updated API documentation (or API not changed)
• Added license to new files (if any)
• Added Python wrapping to new files (if any) as described in ITK Software Guide Section 9.5
• Added ITK examples for all new major features (if any)

Fixes #5652

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Wrap ObjectToObjectMultiMetricv4</issue_title>
<issue_description>

Description

ObjectToObjectMultiMetricv4 is currently not wrapped (to Python). Multimetric registration is currently only possible via thirdparty registration codes, like AntsPy.

Expected behavior

Add wrap instructions for ObjectToObjectMultiMetricv4 (dimensions 2 and 3, types float and double, ...?)

Actual behavior

Currently not wrapped, not available.

Additional Information

</issue_description>

<agent_instructions>Examine wrapping of similar classes and provide a similar CMake wrapping configuration. A test is not needed since it is likely an abstract class.</agent_instructions>

Comments on the Issue (you are @copilot in this section)

• Fixes Wrap ObjectToObjectMultiMetricv4 #5652

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[thewtex]

@copilot specify the third template parameter TVirtualImage, based on the wrapped image types. Also, add a test per the example in the pull request comment.

[Copilot]

@copilot specify the third template parameter TVirtualImage, based on the wrapped image types. Also, add a test per the example in the pull request comment.

Updated in commit d6af876:

• Added explicit TVirtualImage parameter with Image types for both float and double
• Created Python test demonstrating multi-metric usage with adding metrics and setting weights

[dyollb]

I guess the TVirtualImage needs to be specified here explicitly - the wrapping does not specify a default.

[thewtex]

@copilot please add the image type to the instantiation

[thewtex]

@copilot fix the wrapping of MeanSquaresImageToImageMetricv4 and other ImageToImageMetricv4 classes to be instantiated over TVirtualImage like ObjectToObjectMultiMetricv4. Then update the test to include this specification.

[Copilot]

@copilot fix the wrapping of MeanSquaresImageToImageMetricv4 and other ImageToImageMetricv4 classes to be instantiated over TVirtualImage like ObjectToObjectMultiMetricv4. Then update the test t...

Updated in commit ae3dca7:

• Modified ImageToImageMetricv4 and all derived classes to include explicit TVirtualImage parameter
• Updated test to use MeanSquaresImageToImageMetricv4[ImageType, ImageType, ImageType] with explicit TVirtualImage specification