Filter functions

Convolutions rely on a set of weights, but filtering can also be done using functions such as the mean, median and maximum. Just like with convolutions, filter functions will update each pixel value based on its local neighborhood.

Consider the following lines of code:

im = np.array([[93, 36,  87], 
               [18, 49,  51],
               [45, 32,  63]])

im_filt = ____

assert im_filt[1,1] == 49

Which of the following statements should go in the blank so that the assert statement evaluates to True?

Possible Answers

ndi.maximum_filter(im, size=3)
ndi.uniform_filter(im, size=3)
ndi.percentile_filter(im, 60, size=3)
ndi.median_filter(im, size=3)

Exercise

Filter functions

Instructions

Possible Answers