Load an ITK transform file in Python.

gistfile1.py

  def readITKtransform( transform_file ):
    '''
    '''

    # read the transform
    transform = None
    with open( transform_file, 'r' ) as f:
      for line in f:

        # check for Parameters:
        if line.startswith( 'Parameters:' ):
          values = line.split( ': ' )[1].split( ' ' )

          # filter empty spaces and line breaks
          values = [float( e ) for e in values if ( e != '' and e != '\n' )]
          # create the upper left of the matrix
          transform_upper_left = numpy.reshape( values[0:9], ( 3, 3 ) )
          # grab the translation as well
          translation = values[9:]

        # check for FixedParameters:
        if line.startswith( 'FixedParameters:' ):
          values = line.split( ': ' )[1].split( ' ' )

          # filter empty spaces and line breaks
          values = [float( e ) for e in values if ( e != '' and e != '\n' )]
          # setup the center
          center = values

    # compute the offset
    offset = numpy.ones( 4 )
    for i in range( 0, 3 ):
      offset[i] = translation[i] + center[i];
      for j in range( 0, 3 ):
        offset[i] -= transform_upper_left[i][j] * center[i]

    # add the [0, 0, 0] line
    transform = numpy.vstack( ( transform_upper_left, [0, 0, 0] ) )
    # and the [offset, 1] column
    transform = numpy.hstack( ( transform, numpy.reshape( offset, ( 4, 1 ) ) ) )

    return transform