Cfrgtkky

Suppose have an image like this, that consists of two images, 1 and 6 which are kind of superimposed on each other, if you look closely at it.

So, I need to visualize each of these digits separately by having all positive pixels of original image in one image, and all negative pixels in the second one. Is there any way to achieve that in python using matplotlib.pyplot without messing up with the structure of the image? Basically, I need all the white color pixels to be plotted separately from the black color pixels.

You may set all values above or below some threshold to nan, such that they won't appear in the final image.

The following code leaves out the range between 0.4 and 0.6 completely. The yellow background is chosen to show that there are no pixels in that area.

import numpy as np

import matplotlib.pyplot as plt



img = plt.imread("grayscaleimage.png")[:,:,0]



white = np.copy(img)

white[white<0.6] = np.nan



dark = np.copy(img)

dark[dark>0.4] = np.nan



fig = plt.figure()

ax0 = fig.add_subplot(211)

ax1 = fig.add_subplot(223)

ax2 = fig.add_subplot(224)



ax0.imshow(img, vmin=0, vmax=1, cmap="Greys")

ax1.imshow(white, vmin=0, vmax=1, cmap="Greys")

ax2.imshow(dark, vmin=0, vmax=1, cmap="Greys")



for ax in (ax1,ax2):

    ax.set_facecolor("gold")



plt.show()