Saving numpy.ndarray in python as an image

Question

b=ndimage.gaussian_filter(imagefile,5)

Being new to python, not able to figure this out.
how to save b as an image, b is of type 'numpy.ndarray'?

Tried these,
1.

im = Image.fromarray(b)   
im.save("newfile.jpeg")

Error: TypeError("Cannot handle this data type")

2.

imsave('newfile.jpg', b)

Error: ValueError: 'arr' does not have a suitable array shape for any mode.

Which is the right way to save an ndarray into an image?

EDIT:

Solved:

im = Image.fromarray(b)

im.save('newfile.jpeg') worked, The way I was loading the image was wrong,

file = Image.open("abc.jpg")      
imagefile = file.load()

// I was using imagefile after loading, which was not giving proper shape to reconstruct the image.

// Instead If I use file (i.e. directly after opening, I can save by above method)

score 6 · Accepted Answer · 2012-12-11 01:12:02Z

6

I think the best approach is using matplotlib imshow.

using the image library:

import Image
import numpy as np

x = np.array([[1, 2, 3], [4, 5, 6]], np.int32)

im = Image.fromarray(x)
im.save('test.png')

Matplotlib Version:

import numpy as np
import matplotlib.pyplot as plt

x = np.array([[1, 2, 3], [4, 5, 6]], np.int32)
plt.imshow(x) 
plt.savefig("array")

Output of ndarray Hope this helps!

answered Dec 10, 2012 at 23:52

user1786283

Sign up to request clarification or add additional context in comments.

2 Comments

Let me try, I would like to avoid importing few more libraries to just save an array as image.

You can add your previous answer of saving the image, I can save it that way, provided I read the image properly, I have edited the question.