I have the follwoing question. I have a numpy array with size 91x40x21 full of floats and I want to use the median filter. Unfortunatly when i trying to cast it back to an array i get this error:
"cannot reshape array of size 76440 into shape (0,0,0)"
This is my code:
image_view = itk.GetImageViewFromArray(matrix.astype('float32'))
medianImage = itk.MedianImageFilter.New(image_view, Radius = 3)
matrix = itk.GetArrayViewFromImage(medianImage)
If you explicitely update and get output from the image, then everythink will work:
matrix = np.random.random_integers(0,100,(10,6,8))
image_view = itk.GetImageViewFromArray(matrix.astype('Float32'))
medianImageFilter = itk.MedianImageFilter.New(image_view, Radius = 3)
medianImageFilter.Update()
medianImage = medianImageFilter.GetOutput()
newMatrix = itk.GetArrayViewFromImage(medianImage)