CODEWITHSUNDEEP
pythonnumpy
hrithik mahesh
hrithik mahesh

Reputation: 113

need some help to vectorize this code for images

Is there any way to vectorize this code. It is similar to bitwise and? IF "A" is a greyscale image and "B" is a binary image and "C" is matrix of same size of "A" containing zeroes

for row in range(A.shape[0]):
    for col in range(A.shape[1]):
        if B[row, col] == 1:
            C[row, col] = ~A[row, col]
        else:
            C[row, col] = A[row, col]

Upvotes: 1

Views: 65

Answers (2)

zvone
zvone

Reputation: 19382

If A, B, and C are numpy arrays of the same size, you can do operations on the whole array, with approximately the same code you would write for each value.

So, you can do a B == 1 to get a boolean array of the same size as B e.g. [True, False, True, False]. A C[B == 1] will then be a sub-array of C at all indices at which B contains 1.

To sum it all up, you can do this:

C[B == 1] = ~A[B == 1]
C[B != 1] = A[B != 1]

Alternatively, if you don't even have a C to start with:

C = A.copy()
C[B == 1] = ~C[B == 1]

BTW, I would recommend using a, b, and c rather than A, B, and C, because that is how we usually name variables in Python. See PEP-8 for more info.

Upvotes: 1

John Zwinck
John Zwinck

Reputation: 249532

I'd do it using np.where():

C = np.where(B == 1, ~A, A)

Upvotes: 3

Related Questions