I have an image stored in a 2D numpy array. I want to extract all pixel values in a rectangle from that array. The rectangle is defined as ((x1,y1),(x2,y2)) where all x and y s are naturally array indices.
I can extract the pixel values using a nested for loop, but what would be a pythonic way to do this?
Just use slicing. For example:
In [3]: a = numpy.arange(20).reshape((4,5))
In [4]: a
Out[4]:
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19]])
In [5]: a[2:4, 3:5]
Out[5]:
array([[13, 14],
[18, 19]])
In general, you can replace an index with a slice where slices have the form start:stop or, optionally, start:stop:step and variables are allowed:
In [6]: x=2 ; print a[x-1:x+1, :]
[[ 5 6 7 8 9]
[10 11 12 13 14]]
Take a look at numpy's indexing.
import numpy
array = numpy.arange(24).reshape((4, 6))
indices = ((1, 3), (2, 5))
((x1, y1), (x2, y2)) = indices
result = array[x1:x2, y1:y2]
