CODEWITHSUNDEEP
pythonpython-imaging-librarytiff
user3822346
user3822346

Reputation: 31

Python convert 32 bit to 16 bit tiff

Is there a simple way to convert a 32-bit .tif to 16-bit .tif? In IDL, write_tiff allows /short instead of /float. Does Python do that? Instead, I have tried this:

i32 = Image.open('image.tif')
<Image.Image image mode=F size=2016x2016 at 0x102A3E998>
i32.convert('L').save('newimage.tif')

ImageJ (viewer) opens the tif and says that this image is 8-bit, not 16-bit.

Upvotes: 3

Views: 7090

Answers (2)

Antti Haapala -- Слава Україні
Antti Haapala -- Слава Україні

Reputation: 133879

From PIL documentation: The 'L' mode is for 8-bit greyscale images, and your original image is a 32-bit float (mode 'F'), not integer! To get 16 bit integer images, you can try to use 'I;16' for the mode; this is at least supported in my Pillow 2.0.0 (Ubuntu 14.04):

f32 = Image.open('image.tif')
f32.convert('I;16').save('newimage.tif')

Upvotes: 3

Roland Smith
Roland Smith

Reputation: 43495

The Python Imaging Library doesn't support 16-bit pixel depth. See "modes" in the concepts section of the Handbook.

The same goes for the derivative library Pillow.

Try e.g. ImageMagick (which has python bindings) with the -depth option.

Upvotes: 1

Related Questions