[FIXED] Finding the first and last rows that contain a black pixel, in a numpy array

Issue

I have a black and white image of a triangle (it contains only 0 and 255 pixel values).

I've converted it to a numpy array called myArray

Currently, I can find the width of the bottom of the triangle (the number of black pixels) by using this code:

width = (max(numpy.where(myArray == 0)[1])) - (min(numpy.where(myArray == 0)[1]))

If the triangle was flipped upside-down, width would then apply to the top of the upside-down triangle.

What i'm trying to do is determine if the triangle is pointing up or down.

I could do this by finding the first row that contains a black pixel, and counting the number of black pixels in that row, calling this firstRow

and finding the last row that contains a black pixel, and counting the number of black pixels in that row, calling that lastRow

Then, if firstRow < lastRow, the triangle is pointing up.

What is the best way to calculate firstRow and lastRow?

Solution

With myArray with 255 for black pixel such as

array([[  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., 255.,   0.,   0.,   0.,   0.],
       [  0.,   0.,   0., 255., 255., 255.,   0.,   0.,   0.],
       [  0.,   0., 255., 255., 255., 255., 255.,   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.]])

you can get all the rows with at least one black pixel with np.where after using sum on axis=1:

print (np.where(myArray.sum(axis=1)))
(array([3, 4, 5], dtype=int64),)

If you want to get the row with the maximum number of black pixels, you can use np.argmax still after sum on axis=1:

print (np.argmax(myArray.sum(axis=1)))
5

To know if the triangle is up or down, one way is to check if the argmax is the np.max element in the np.where(myArray.sum(axis=1)), then it would be up.

myArray_sum = myArray.sum(axis=1)
if np.max(np.where(myArray_sum)) == np.argmax(myArray_sum):
    print ('up')
else:
    print ('down')

If you want the first and last row, here is one way but it is related to the value of the black pixel.

myArray_sum = myArray.sum(axis=1)
firstRow = np.argmax(myArray_sum == 255)
lastRow = np.argmax(myArray_sum)

Answered By - Ben.T