i am playing around with scikit image restoration package and successfully ran the unsupervised_wiener algorithm on some made up data. In this simple example it does what I expect, but on my more complicated dataset it returns a striped pattern with extreme values of -1 and 1.
I would like to fiddle with the parameters to better understand what is going on, but I get the error as stated in the question. I tried scikit image version 0.19.3 and downgraded to scikit image version 0.19.2, but the error remains.
The same goes for the "other parameters":https://scikit-image.org/docs/0.19.x/api/skimage.restoration.html#skimage.restoration.unsupervised_wiener
Can someone explain why I can't input parameters?
The example below contains a "scan" and a "point-spread-function". I convolve the scan with the point spread function and then reverse the process using the unsupervised wiener deconvolution.
import numpy as np
import matplotlib.pyplot as plt
from skimage import color, data, restoration
import pickle
rng = np.random.default_rng()
from scipy.signal import convolve2d as conv2
scan = np.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, 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, 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, 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, 0.5, 0.5, 0.5, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0.5, 1, 0.5, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0.5, 1, 0.5, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0.5, 1, 0.5, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0.5, 0.5, 0.5, 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, 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, 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, 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]
])
print(scan.shape)
psf = np.array([
[1, 1, 1, 1, 1],#1
[1, 0, 0, 0, 1],#2
[1, 0, 0, 0, 1],#3
[1, 0, 0, 0, 1],#4
[1, 1, 1, 1, 1]#5
])
psf = psf/(np.sum(psf))
print(psf)
scan_conv = conv2(scan, psf, 'same')
deconvolved1, _ = restoration.unsupervised_wiener(scan_conv, psf, max_num_iter=10)
fig, ax = plt.subplots(nrows=1, ncols=3, figsize=(8, 5),
sharex=True, sharey=True)
ax[0].imshow(scan, vmin=scan.min(), vmax=1)
ax[0].axis('off')
ax[0].set_title('Data')
ax[1].imshow(scan_conv)
ax[1].axis('off')
ax[1].set_title('Data_distorted')
ax[2].imshow(deconvolved1)
ax[2].axis('off')
ax[2].set_title('restoration1')
fig.tight_layout()
plt.show()
