Hello, I’m new to SimpleITK. My goal is to registrate two 2D image and merge them into a bigger one. I noticed that it is similar to the function of Tutorial notebook7, the only difference is that the fixed image is not under the moving image but at the left side of it.
I modified the code to this:
minimal_overlap = np.array(
moving_image.TransformContinuousIndexToPhysicalPoint(
(moving_image.GetWidth() - self._min_column_overlap, -self._row_overlap)
)
) - np.array(fixed_image.GetOrigin())
maximal_overlap = np.array(
moving_image.TransformContinuousIndexToPhysicalPoint(
(moving_image.GetWidth() - self._max_column_overlap, self._row_overlap)
)
) - np.array(fixed_image.GetOrigin())
and
metric_sampling_percentage = 0.5
row_overlap = 0.2 * left_image.GetHeight()
min_column_overlap = 20
max_column_overlap = 0.4 * left_image.GetWidth()
dx_step_num = 15
dy_step_num = 15
But there are just no valid candidate.
This is the sample image.