我有形状为(112, 176, 112)
的 nifti 文件 (.nii(。我想为它添加另一个维度,以便它变得(112, 176, 112, 3)
.当我尝试img2 = np.arange(img).reshape(112,176,112,3)
时,我收到一个错误。 是否可以使用np.reshape
或np.arange
或任何其他方式来做到这一点?
法典:
import numpy as np
import nibabel as nib
filepath = 'test.nii'
img = nib.load(filepath)
img = img.get_fdata()
img = np.arange(img).reshape(112,176,112,3)
img = nib.Nifti1Image(img, np.eye(4))
img.get_data_dtype() == np.dtype(np.int16)
img.header.get_xyzt_units()
nib.save(img, 'test_add_channel.nii')
错误:
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-16-f6f2a2d91a5d> in <module>
8 print(img.shape)
9
---> 10 img2 = np.arange(img).reshape(112,176,112,3)
11
12 img = nib.Nifti1Image(img, np.eye(4))
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()
你可以这样做:
import numpy as np
img = np.random.rand(112, 176, 112) # Your image
new_img = img.reshape((112, 176, 112, -1)) # Shape: (112, 176, 112, 1)
new_img = np.concatenate([new_img, new_img, new_img], axis=3) # Shape: (112, 176, 112, 3)
可能这是其他更好的方法,但上面的代码为您提供了所需的输出。