Converting image to numpy array in python

Question

I'm using Python + Scipy + Scikit-image + numpy for the first time.

I'm using this page for help, and I've only changed the given code a bit to pick up an image of my liking:

tree = misc.imread('C:\\Users\\app\\Pictures\\treephoto1.jpg')
type(tree)
<type 'numpy.ndarray' >
tree.shape, tree.dtype((512, 512), dtype('uint8'))

But I'm getting the following error:

 type(tree) <type 'numpy.ndarray'>
                                   ^
SyntaxError: invalid syntax

What's wrong with the syntax? I'm using python 2.7 on Windows, and all the related toolkits are also according to Python 2.7.

I need to convert the image to a 2-D numpy array so that I can use the canny edge detector with it.

twil · Accepted Answer · 2014-04-04 10:11:54Z

3

Writing without thinking may be dangerous but in your case it is just wrong. The page you've mentioned shows this:

>>> lena = misc.imread('lena.png')
>>> type(lena)
<type 'numpy.ndarray'>
>>> lena.shape, lena.dtype
((512, 512), dtype('uint8'))

>>> is Python's interactive console prompt string. It is there were commands go.

<type 'numpy.ndarray'> and ((512, 512), dtype('uint8')) are results of commands. So your corresponding code should be only

tree = misc.imread('C:\\Users\\app\\Pictures\\treephoto1.jpg')
type(tree)
tree.shape, tree.dtype

Notice that

type(tree)
tree.shape, tree.dtype

do nothing and just show you information about your image data.

Update

Basically (not always) your image is layered. RGB is three separate layers. So if your filtering isn't aware of it you'll have to separate layers yourself.

Hope that helps.

edited Apr 4, 2014 at 10:11

answered Apr 3, 2014 at 18:02

twil

6,1721 gold badge34 silver badges31 bronze badges

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

2 Comments

user961627 Over a year ago

I see! Thanks. Okay I think I missed out an important word in the question, I'm editing it now. I want the numpy array to be 2D. I learned that images are by default 2-D arrays, but when I run "edges = filter.canny(tree)", it gives me an error saying that the input image must be a 2-D array.

user961627 Over a year ago

Alright I just found out! I can convert it to a grayscale image to turn it into a 2D array. I did this: graytree = color.rgb2gray(tree). Just a side question: are most image processing algorithms carried out on grayed images?

Collectives™ on Stack Overflow

Converting image to numpy array in python

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related