Applying a Fourier Transform to an image

An image is Fourier-transformed using the alg.FFT() algorithm object:

#imports
from ost.io import LoadImage
from ost.img.alg import FFT
from iplt.gui import Viewer
#Load the image
im=LoadImage("FILENAME.tif") # load the image
# create an instance of the fft algorithm object
fft=FFT() 
# do the actual Fourier transformation
im_ft=im.Apply(fft) 
# back-transform
im2 = im_ft.Apply(fft) 
# if this is run from within giplt, open viewers to look at the images
Viewer(im)
Viewer(im_ft)
Viewer(im2)

It is not really necessary to create the fft instance, a temporary object can be used, since the alg.FFT algorithm object is stateless. In addition, the algorithm can be applied in-place to avoid the creation of a second image:

#imports
from ost.io import LoadImage
from ost.img.alg import FFT
#Load the image
im=LoadImage("FILENAME.tif") # load the image
# do the actual Fourier transformation, in-place using temporary object
im.ApplyIP(FFT()) 
# repeating this command will do the back-transform
im.ApplyIP(FFT()) 

Please note that the FFT() algorithm does not require a direction to be given, this is implicitly determined by the underlying image state: a SPATIAL image will always be transformed to the FREQUENCY domain, and vice-versa.

---

back to Tutorials