How to mask an image gray scale using numpy array slicing

Question

I need to replace 8 bits values (0 to 255) indexed set of an image (final image), following a "map values" from another image (second image) gray scale which related map indexes was chosen from a primary image.

In fact this is similar thing that MATLAB does with

 indexS =  find(image1 == integer ('could be a integer from 1 to 255')) 
 imagfinal(indexS) = imagsecondary(indexS).

I tried following examples for python/matlab find() on stack, for ex.: MATLAB-style find() function in Python. And the related ones...

I tried n.nonzero , np.argwhere and np.where, but I am really confuse.

I have three source images, let's say A, B, C, same shape, eg. (100x100) with diverse 0 to 255 values, I mean they are completely gray scales different each other.

So, 1st step - I need to get all indexes with values = 1 (but could be , 10, 196 , up to 255) from A, so I did:

Aboolean = np.equal(A,1)

result is

       [False, False, False, ..., False, False, False],
       [False, False, False, ..., False, False, False],...

then I tried use those boolean index array results for getting the values from B:

Bnew = B[Aboolean]

But it did not work for furthers steps, because the result is a map of values and the indexes are lost...

The values from Bnew are supposed to substitute the respective values 8-bits on C image,I mean those 8-bits values into the same position (or same indexes), remembering that B and C (also A) have the same shape/size array (100x100).

So I tried again:


D = np.where(Aboolean,B,C)

when plotting the image, the final result is just the same image C !! No modifications, at all.


fig, ax = plt.subplots(nrows=1, figsize=(16,20))
ax.imshow(D, cmap='gray',interpolation='nearest')

results same image 'C'

My goal is a kind of replacing a set of values from B upon C (ruled by same index positions) , that was sliced following a map of indexes of conditions upon A.

locometro · Accepted Answer · 2019-06-21 02:06:23Z

1

You can do this by using the boolean indexing from A to directly copy the values from C into B (if you don't want to modify the original B, first create a copy using B.copy()).

>>> import numpy as np
>>> A = np.array([0,0,1,0,0])
>>> B = np.array([1,2,3,4,5])
>>> C = np.array([10,9,8,7,6])
>>> B[A==1] = C[A==1]

>>> B
array([1, 2, 8, 4, 5])

EDIT:

C[A==1] = B[A==1]

edited Jun 21, 2019 at 2:06

locometro

1053 bronze badges

answered Jun 21, 2019 at 1:01

Tom Holmes

1,20410 silver badges11 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

locometro Over a year ago

Hi Tom, I mean copy values from B into C. But the logical seems clear now.

locometro Over a year ago

What about the non-zero values? The effect of avoiding non-zero values, as is done by function find() in matlab. In python, would be : C[ np.nonzero(cond) ] = B[ np.nonzero( cond ) ] ?

Tom Holmes Over a year ago

not sure I understand your question. what do you want to do with the non-zero values?

locometro Over a year ago

I am trying replicate the same logical, if I use matlabcode: find()... Imagfinal(index) = Imag2(index). As explained in: stackoverflow.com/questions/14388786/…

Tom Holmes Over a year ago

Ah I see. np.nonzero filters the array down to non-zero elements, as opposed to creating an index of non-zero elements. What I think you want can be done the same way as the above solution: C[A != 0] = B[A != 0]

|

Collectives™ on Stack Overflow

How to mask an image gray scale using numpy array slicing

1 Answer 1

6 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

6 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related