Assigning values ​​based on a small mask placed above a pixel

Question

Problem

I have a circular boolean mask of arbitrary radius (always perfectly symmetrical):

array([[False, False,  True, False, False],
       [False,  True,  True,  True, False],
       [ True,  True,  True,  True,  True],
       [False,  True,  True,  True, False],
       [False, False,  True, False, False]], dtype=bool)

Then I have a big matrix of uint8, imageand a pair in the matrix, which can be any valid index point.

What I want to do is apply this mask to this point in the image so that I can essentially just put the circle at that point in image.

This is pretty easy to do in the middle of the image. You can simply:

image[point[0] - radius:point[0] + radius + 1, point[1] - radius:point[1] + radius + 1] = circle_mask

But naturally, this does not handle border validation, which in this case seems rather complicated, because I have to make sure that the range imagethat we assign is the same size as the assigned mask.

Example

point - (1, 1), 2, , , image , :

array([[ 1.,  1.,  1.,  0.,  0.,  ... ,  0.],
       [ 1.,  1.,  1.,  1.,  0.,  ... ,  0.],
       [ 1.,  1.,  1.,  0.,  0.,  ... ,  0.],
       [ 0.,  1.,  0.,  0.,  0.,  ... ,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  ... ,  0.],
       ...,
       [ 0.,  0.,  0.,  0.,  0.,  ... ,  0.]])

, , :

# Initialization of stuff so this is runnable
point = (1, 1)
radius = 2
image = np.zeros((10, 10))
x, y = np.ogrid[-radius : radius + 1, -radius : radius + 1]
circle_mask = x**2 + y**2 <= radius**2

# My solution to the problem
image_min_row = max(point[0] - radius, 0)
image_min_col = max(point[1] - radius, 0)
image_max_row = min(point[0] + radius + 1, image.shape[0])
image_max_col = min(point[1] + radius + 1, image.shape[1])

mask_min_row = max(radius - point[0], 0)
mask_min_col = max(radius - point[1], 0)
mask_max_row = min(image.shape[0] - point[0] + radius, circle_mask.shape[0])
mask_max_col = min(image.shape[1] - point[1] + radius, circle_mask.shape[1])

temp_mask = circle_mask[mask_min_row:mask_max_row, mask_min_col:mask_max_col]
image[image_min_row:image_max_row, image_min_col:image_max_col][temp_mask] = 1

. , , . , . -, , , point , .

NumPy ?

Answer 1

, :

x = np.arange(image.shape[0]) 
y = np.arange(image.shape[1])
image[np.add.outer((x-point[0])**2, (y-point[1])**2) <= radius**2] = 1

x, y - image, , , .

Answer 2

Scipy binary-dilation -

from scipy.ndimage.morphology import binary_dilation as dilate

mask_mapped = np.zeros(image.shape,dtype=bool)
mask_mapped[points[...,0],points[...,1]] = 1
image[dilate(mask_mapped,circle_mask)] = 1 

№1: -

In [88]: points = np.array([1,1])

In [89]: image = np.zeros((10, 10))

In [90]: mask_mapped = np.zeros(image.shape,dtype=bool)
    ...: mask_mapped[points[...,0],points[...,1]] = 1
    ...: image[dilate(mask_mapped,circle_mask)] = 1
    ...: 

In [91]: print image
[[ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  1.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]]

№2: -

In [96]: points = np.array([[1,1],[7,3],[8,9]])

In [97]: image = np.zeros((10, 10))

In [98]: mask_mapped = np.zeros(image.shape,dtype=bool)
    ...: mask_mapped[points[...,0],points[...,1]] = 1
    ...: image[dilate(mask_mapped,circle_mask)] = 1
    ...: 

In [99]: print image
[[ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  1.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  1.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  1.  1.  1.  0.  0.  0.  0.  1.]
 [ 0.  1.  1.  1.  1.  1.  0.  0.  1.  1.]
 [ 0.  0.  1.  1.  1.  0.  0.  1.  1.  1.]
 [ 0.  0.  0.  1.  0.  0.  0.  0.  1.  1.]]

zeros , , :

In [102]: print dilate(mask_mapped,circle_mask).astype(float)
[[ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  1.  0.  0.  0.  0.  0.  0.]
 [ 1.  1.  1.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  1.  0.  0.  0.  0.  0.  0.]
 [ 0.  0.  1.  1.  1.  0.  0.  0.  0.  1.]
 [ 0.  1.  1.  1.  1.  1.  0.  0.  1.  1.]
 [ 0.  0.  1.  1.  1.  0.  0.  1.  1.  1.]
 [ 0.  0.  0.  1.  0.  0.  0.  0.  1.  1.]]