This demonstration illustrates the decomposition of a step function image into cossenoidal waves of increasing frequencies.
The image is created using.
>>> import Numeric
>>> f = Numeric.ones((128, 128)) * 50
>>> x, y = [], map(lambda k:k%128, range(-32,32))
>>> for i in range(128): x = x + (len(y) * [i])
>>> y = 128 * y
>>> Numeric.put(f, iasub2ind([128,128], x, y), 200)
>>> iashow(f)
(128, 128) Min= 50 Max= 200 Mean=125.000 Std=75.00
The DFT is computed and displayed
>>> F = iadft(f)
>>> E = iadftview(F)
>>> iashow(E)
(128, 128) Min= 0 Max= 255 Mean=0.737 Std=11.72
>>> g,d = iaplot(iafftshift(F[:,0]).real)
>>> g('set data style impulses')
>>> g.plot(d)
>>>