For a program I'm writing, I need to convert an RGB image to grayscale and read it as a NumPy array using PIL.
But when I run the following code, it converts the image not to grayscale, but to a strange color distortion a bit like the output of a thermal camera, as presented.
Any idea what the problem might be?
Thank you!
http://www.loadthegame.com/wp-content/uploads/2014/09/thermal-camera.png
from PIL import Image
from numpy import *
from pylab import *
im = array(Image.open('happygoat.jpg').convert("L"))
inverted = Image.fromarray(im)
imshow(inverted)
show()
matplotlib's imshow is aimed at scientific representation of data - not just image data. By default it's configured to use a high constrast color palette.
You can force it to display data using grayscale by passing the following option:
import matplotlib.cm
imshow(inverted, cmap=matplotlib.cm.Greys_r)
Add this code to view/display an image:
from PIL import Image;
from numpy import *
from pylab import *
im = array(Image.open('happygoat.jpg').convert("L"));
inverted = Image.fromarray(im);
inverted
Related
I have 2 modules in my project: first works with image in bytes format, second requires skimage object. I need to combine them.
I have this code:
import io
from PIL import Image
import skimage.io
area = (...)
image = Image.open(io.BytesIO(image_bytes))
image = Image.crop(area)
image = skimage.io.imread(image)
But i get this error:
How can i convert an image (object/variable) to skimage? I don't necessarily need PIL Image, this is just one way to work with bytes image, cause i need to crop my image
Thanks!
Scikit-image works with images stored as Numpy arrays - same as OpenCV and wand. So, if you have a PIL Image, you can make a Numpy array for scikit-image like this:
# Make Numpy array for scikit-image from "PIL Image"
na = np.array(YourPILImage)
Just in case you want to go the other way, and make a PIL Image from a Numpy array, you can do:
# Make "PIL Image" from Numpy array
pi = Image.fromarray(na)
I'm reading in an image with OpenCV, and trying to do something with it in numpy (rotate 90deg). Viewing the result with imshow from matplotlib, it all seems to be working just fine - image is rotated. I can't use drawing methods from OpenCV on the new image, however. In the following code (I'm running this in a sagemath cloud worksheet):
%python
import cv2
import matplotlib.pyplot as plt
import numpy as np
import os, sys
image = np.array( cv2.imread('imagename.png') )
plt.imshow(image,cmap='gray')
image = np.array(np.rot90(image,3) ) # put it right side up
plt.imshow(image,cmap='gray')
cv2.rectangle(image,(0,0),(100,100),(255,0,0),2)
plt.imshow(image,cmap='gray')
I get the following error on the cv2.rectangle() command:
TypeError: Layout of the output array img is incompatible with cv::Mat (step[ndims-1] != elemsize or step[1] != elemsize*nchannels)
The error goes away if I use np.array(np.rot90(image,4) ) instead (i.e. rotate it 360). So it appears that the change in dimensions is messing it up. Does OpenCV store the dimensions somewhere internally that I need to update or something?
EDIT: Adding image = image.copy() after rot90() solved the problem. See rayryeng's answer below.
This is apparently a bug in the Python OpenCV wrapper. If you look at this question here: np.rot90() corrupts an opencv image, apparently doing a rotation that doesn't result back in the original dimensions corrupts the image and the OP in that post experiences the same error you are having. FWIW, I also experienced the same bug.... no idea why.
A way around this is to make a copy of the image after you rotate, and then show the image. This I can't really explain, but it seems to work. Also, make sure you call plt.show() at the end of your code to show the image:
import cv2
import matplotlib.pyplot as plt
import numpy as np
import os, sys
image = np.array( cv2.imread('imagename.png') )
plt.imshow(image,cmap='gray')
image = np.array(np.rot90(image,3) ) # put it right side up
image = image.copy() # Change
plt.imshow(image,cmap='gray')
cv2.rectangle(image,(0,0),(100,100),(255,0,0),2)
plt.imshow(image,cmap='gray')
plt.show() # Show image
I faced the same problem with numpy 1.11.2 and opencv 3.3.0. Not sure why, but this did the job for me.
Before using cv2.rectangle, add the line below:
image1 = image1.transpose((1,0)).astype(np.uint8).copy()
Reference
Convert data type works for my problem.
The image is of type np.int64 before the convert.
image = image.astype(np.int32) # convert data type
I'm working on an python program to display images of stars. The images are 16-bit grayscale tiffs.
If I try to display them in an extern program, e.g. ImageMagick they are correct but if I load them in python and then use 'show()' or implement them in a canvas in Tkinter they are, unless a few pixel, totally white.
So I estimate python sets every pixel above 255 to white but I don't know why. If I load the image and then save it as tiff again, ImageMagick can show it correct.
Thanks for help.
Try to convert the image to a numpy array and display that:
import Image
import matplotlib.pyplot as plt
import numpy as np
img = Image.open('image.tiff')
arr = np.asarray(img.getdata()).reshape(img.size[1], img.size[0])
plt.imshow(arr)
plt.show()
You can change the color mapping too:
from matplotlib import cm
plt.imshow(arr, cmap=cm.gray)
I'm trying to display a PNG file using matplotlib and of course, python. For this test, I've generated the following image:
Now, I load and transform the image into a multidimensional numpy matrix:
import numpy as np
import cv2
from matplotlib import pyplot as plt
cube = cv2.imread('Graphics/Display.png')
plt.imshow(cube)
plt.ion()
When I try to plot that image in matplotlib, the colors are inverted:
If the matrix does not have any modifications, why the colors in the plot are wrong?
Thanks in advance.
It appears that you may somehow have RGB switched with BGR. Notice that your greens are retained but all the blues turned to red. If cube has shape (M,N,3), try swapping cube[:,:,0] with cube[:,:,2]. You can do that with numpy like so:
rgb = numpy.fliplr(cube.reshape(-1,3)).reshape(cube.shape)
From the OpenCV documentation:
Note: In the case of color images, the decoded images will have the
channels stored in B G R order.
Try:
plt.imshow(cv2.cvtColor(cube, cv2.COLOR_BGR2RGB))
As others have pointed out, the problem is that numpy arrays are in BGR format, but matplotlib expects the arrays to be ordered in a different way.
You are looking for scipy.misc.toimage:
import scipy.misc
rgb = scipy.misc.toimage(cube)
Alternatively, you can use scipy.misc.imshow().
Color image loaded by OpenCV is in BGR mode. However, Matplotlib displays in RGB mode.
So we need to convert the image from BGR to RGB:
plt.imshow(cv2.cvtColor(cube, cv2.COLOR_BGR2RGB))
I am looking for a way to rescale the matrix given by reading in a png file using the matplotlib routine imread,
e.g.
from pylab import imread, imshow, gray, mean
from matplotlib.pyplot import show
a = imread('spiral.png')
#generates a RGB image, so do
show()
but actually I want to manually specify the dimension of $a$, say 200x200 entries, so I need some magic command (which I assume exists but cannot be found by myself) to interpolate the matrix.
Thanks for any useful comments : )
Cheers
You could try using the PIL (Image) module instead, together with numpy. Open and resize the image using Image then convert to array using numpy. Then display the image using pylab.
import pylab as pl
import numpy as np
from PIL import Image
path = r'\path\to\image\file.jpg'
img = Image.open(path)
img.resize((200,200))
a = np.asarray(img)
pl.imshow(a)
pl.show()
Hope this helps.