裁剪中心部分的numpy图像

Question

假设我有一个宽x和高y的numpy图像.我必须将图像的中心部分裁剪为宽度cropx和height cropy.让我们假设cropx和cropy是正非零整数并且小于相应的图像大小.将切片应用于输出图像的最佳方法是什么？

Answer 1

沿着这些方向的东西 -

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]

样品运行 -

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

基于@Divakar答案的更通用的解决方案:

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]

如果我们有一个数组 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]])

我们可以剪辑它cropND(a, (5,5)),你会得到:

>>> 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]])

它不仅适用于2D图像,也适用于3D图像.

祝你今天愉快.