How can I prevent Matplotlib from autoscaling images to the same size in subplots?

I’m trying to display three images of different sizes side-by-side using Matplotlib. I want each subplot to maintain the original size of the images without auto scaling them to be the same size.

img_512 = down_sample(1024, 512, img)

img_512_upsample = up_sample(512, 1024, img_512)

fig, ax = plt.subplots(1,3, figsize=(15,10))

ax[0].imshow(img, cmap='gray', aspect='equal', extent=[0, img.shape[1], img.shape[0], 0])
ax[0].set_title(f"Original Image")
ax[0].axis('on')

ax[1].imshow(img_512, cmap='gray', aspect='equal', extent=[0, img_512.shape[1], img_512.shape[0], 0])
ax[1].set_title(f"Down Sample to 512x512")
ax[1].axis('on')


ax[2].imshow(img_512_upsample, cmap='gray', aspect='equal', extent=[0, img_512_upsample.shape[1], img_512_upsample.shape[0], 0])
ax[2].set_title(f"Up Sample to 1024x1024")
ax[2].axis('on')

plt.autoscale(False)
plt.tight_layout()
plt.show()

Image of my result

I tried turning off auto scaling, setting aspect to equal and tried setting the extent to the sizes of the images.

If you want the subplots the same size, use sharex/sharey in the call to subplots:

import matplotlib.pyplot as plt
import numpy as np

img = np.zeros((1024, 1024))
img_512 = np.zeros((512, 512))
img_512_upsample = np.zeros((1024, 1024))

fig, ax_arr = plt.subplots(1,3, figsize=(15,10), sharex=True, sharey=True)

ax_arr[0].imshow(img, cmap='gray', aspect='equal', extent=[0, img.shape[1], img.shape[0], 0])
ax_arr[0].set_title(f"Original Image")

ax_arr[1].imshow(img_512, cmap='gray', aspect='equal', extent=[0, img_512.shape[1], img_512.shape[0], 0])
ax_arr[1].set_title(f"Down Sample to 512x512")

ax_arr[2].imshow(img_512_upsample, cmap='gray', aspect='equal', extent=[0, img_512_upsample.shape[1], img_512_upsample.shape[0], 0])
ax_arr[2].set_title(f"Up Sample to 1024x1024")

plt.show()

If instead you want the middle subplot half the height and width, use width_ratios in the call to subplots:

fig, ax_arr = plt.subplots(1,3, figsize=(15,10), width_ratios=[2, 1, 2])

You can align the subplots by their top or bottom edge using the anchor property - "N" for north or "S" for south:

for ax in ax_arr:
    ax.set_anchor("N")