0x64746b / python-cpbd Goto Github PK

I am unable to read an image for cpbd.compute() to process. The example usage in README is broken.

This code:
from scipy import ndimage
input_image = ndimage.imread('/tmp/LIVE_Images_GBlur/img4.bmp', mode='L')

Results in:
AttributeError: module 'scipy.ndimage' has no attribute 'imread'

The docs for Scipy say imread was removed in version 1.2,
"imread is deprecated in SciPy 1.0. 0, and will be removed in 1.2."
https://docs.scipy.org/doc/scipy-1.2.1/reference/generated/scipy.misc.imread.html

I have found a StackOverflow page that recommends using image.io:
https://stackoverflow.com/questions/15345790/scipy-misc-module-has-no-attribute-imread

However, there is no mode="L" parameter, just a format param, which overrides the image format guess from the extension (ie format="JPEG").

 import imageio
 iio_image = imageio.imread(inputpath)
 cpbd_focus1 = cpbd.compute(iio_image)

However, using plain imageio.imread results in cpbd complaining about the input shape, because its in RGB.
ValueError: The parameter image must be a 2-dimensional array

If I convert the image to grayscale, it works.

import cpbd
import cv2
import imageio
iio_image = imageio.imread('testimg.jpg')
grey_iio_image = cv2.cvtColor(iio_image, cv2.COLOR_BGR2GRAY)
cpbd.compute(grey_iio_image)
0.25540727105384264

Please update your README to address this change in modern scipy.

Thank you!!

I would like to know how to compute CPBD score on diffusion MRI 4-D data, since the default function works on 2D images.

Moreover, I wanted to make sure I am applying the commands correctly.
Diffusion MRI data are 4-D grayscale images in nifti (.nii.gz) format.

Here is the code I use -for now on a single slice of a diffusion MRI volume in order to have a 2-D image:

            # select single slice and volume from 4-D diffusion MRI data to have 2-D one
             orig = data[:, :, sli, vol]
            matplotlib.image.imsave('blur1_rev.png', orig)
             input_image1 = cv2.imread('blur1_rev.png')
             input_image1 = cv2.cvtColor(input_image1, cv2.COLOR_BGR2GRAY)
             cpbd.compute(input_image1)

thank you in advance ,
Rosella