How can I read single-channel 32-bit integer images with python?

Question

I want to read single-channel 32-bit integer image saved as ssv file. I have tried so far(see the following code) but without much success.

Here is the code

Please let me know if you have any idea of what is missing?

Answer 1

You as user should have the necessary information about this image-data which is needed. I can only assume things and work on that.

Opening the file it seems it's:

• text-based (supported by your ImageJ-comment)
• using uint32 (numbers too big for uint8; the classic case)

Follwing above and ImageJ's docs:

Opens a tab-delimited text file as a 32-bit real image

i would do:

import numpy as np
import matplotlib.pyplot as plt                      # just for demo

img_raw = np.loadtxt('test.ssv', dtype=float)        # casting-early to float
img_float_01 = img_raw / 4294967295.  # max uint32   # normalize to float in [0, 1]

plt.imshow(img_raw, cmap='gray')
plt.show()

which outputs:

So it seems we successfully read that image into a numpy-array.

You need now to:

• think what steps you need to use that array in QT
• think if you will use above code or look for some more safe route like using scikit-image for example (if you need [0,1] normalization; reading will always be done by numpy's loadtxt)