cv2.imwrite and pyplot.imshow displaying grey scale images differently

Question

I have a bunch of medical images in a hdf5 file.

 path_to_data = "data/train_large_axial_350_x_350.hdf5"
 data = h5py.File(path_to_data,"r")
 image_list = data["images"]
 image = image_list[5]

If I do

 cv2.imwrite("maps/image.jpg", image)

I get:

But, if I do

plt.imshow(image)

I get:

What is the cause of this and how can I plot the image so that it looks like that which is produced by cv2.imwrite?

Does this answer your question? [why cv2.imwrite() changes the color of pics?](https://stackoverflow.com/questions/42406338/why-cv2-imwrite-changes-the-color-of-pics) — abhilb, Dec 04 '19 at 12:32
@abhilb It's a grayscale image -- even if it's 3-channel, alll the channels are identical, so the order is irrelevant in this case. — Dan Mašek, Dec 04 '19 at 15:05
Maybe you have a default matplotlib colormap active. What is the result for `plt.imshow(img, cmap='gray')`? — Stef, Dec 04 '19 at 16:08
plt.imshow(image_list[0], cmap='gray') produces the same result — GhostRider, Dec 06 '19 at 09:16

score 0 · Answer 1 · answered Dec 06 '19 at 14:50

This is because matplotlib plots the image in the range of the gray values of your image. For example, if your image contains gray values between 21 to 88, then matplotlib plots it with this range, and not between 0 and 255.

This can be fixed by the following:

# from matplotlib import cm    
plt.imshow(image, cmap=cm.gray, vmin=0, vmax=255)

For a more in-depth discussion, check out this github issue.

cv2.imwrite and pyplot.imshow displaying grey scale images differently

1 Answers1