Part of the center in the middle of the numpy image

Question

Let's say I have a numerical image of some width x and height y. I have to crop the center of the image to cropx width and height. Suppose cropx and yield are positive integers other than zero and smaller than the corresponding image size. What is the best way to apply slicing to the output image?

Answer 1

Something in this direction -

def crop_center(img,cropx,cropy): y,x = img.shape startx = x//2-(cropx//2) starty = y//2-(cropy//2) return img[starty:starty+cropy,startx:startx+cropx] 

Run Example -

 In [45]: img Out[45]: array([[88, 93, 42, 25, 36, 14, 59, 46, 77, 13, 52, 58], [43, 47, 40, 48, 23, 74, 12, 33, 58, 93, 87, 87], [54, 75, 79, 21, 15, 44, 51, 68, 28, 94, 78, 48], [57, 46, 14, 98, 43, 76, 86, 56, 86, 88, 96, 49], [52, 83, 13, 18, 40, 33, 11, 87, 38, 74, 23, 88], [81, 28, 86, 89, 16, 28, 66, 67, 80, 23, 95, 98], [46, 30, 18, 31, 73, 15, 90, 77, 71, 57, 61, 78], [33, 58, 20, 11, 80, 25, 96, 80, 27, 40, 66, 92], [13, 59, 77, 53, 91, 16, 47, 79, 33, 78, 25, 66], [22, 80, 40, 24, 17, 85, 20, 70, 81, 68, 50, 80]]) In [46]: crop_center(img,4,6) Out[46]: array([[15, 44, 51, 68], [43, 76, 86, 56], [40, 33, 11, 87], [16, 28, 66, 67], [73, 15, 90, 77], [80, 25, 96, 80]]) 

Answer 2

More general solution based on @Divakar answer:

 def cropND(img, bounding): start = tuple(map(lambda a, da: a//2-da//2, img.shape, bounding)) end = tuple(map(operator.add, start, bounding)) slices = tuple(map(slice, start, end)) return img[slices] 

and if we have an array a

 >>> a = np.arange(100).reshape((10,10)) array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9], [10, 11, 12, 13, 14, 15, 16, 17, 18, 19], [20, 21, 22, 23, 24, 25, 26, 27, 28, 29], [30, 31, 32, 33, 34, 35, 36, 37, 38, 39], [40, 41, 42, 43, 44, 45, 46, 47, 48, 49], [50, 51, 52, 53, 54, 55, 56, 57, 58, 59], [60, 61, 62, 63, 64, 65, 66, 67, 68, 69], [70, 71, 72, 73, 74, 75, 76, 77, 78, 79], [80, 81, 82, 83, 84, 85, 86, 87, 88, 89], [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]]) 

We can crop it with cropND(a, (5,5)) , you will get:

 >>> cropND(a, (5,5)) array([[33, 34, 35, 36, 37], [43, 44, 45, 46, 47], [53, 54, 55, 56, 57], [63, 64, 65, 66, 67], [73, 74, 75, 76, 77]]) 

It works not only with 2D images, but also with 3D images.

Have a nice day.

Answer 3

Thank you, Divacar.

Your answer made me go in the right direction. I came up with this using negative slice offsets to count from the end:

 def cropimread(crop, xcrop, ycrop, fn): "Function to crop center of an image file" img_pre= msc.imread(fn) if crop: ysize, xsize, chan = img_pre.shape xoff = (xsize - xcrop) // 2 yoff = (ysize - ycrop) // 2 img= img_pre[yoff:-yoff,xoff:-xoff] else: img= img_pre return img 

Answer 4

A simple modification to @Divakar's answer, which saves the image channel:

  def crop_center(self, img, cropx, cropy): _, y, x = img.shape startx = x // 2 - (cropx // 2) starty = y // 2 - (cropy // 2) return img[:, starty:starty + cropy, startx:startx + cropx]