2.6. Image manipulation and processing using NumPy and SciPy¶

Authors: Emmanuelle Gouillart, Gaël Varoquaux

This section addresses basic image manipulation and processing using the core scientific modules NumPy and SciPy. Some of the operations covered by this tutorial may be useful for other kinds of multidimensional array processing than image processing. In particular, the submodule scipy.ndimage provides functions operating on n-dimensional NumPy arrays.

See also

For more advanced image processing and image-specific routines, see the tutorial scikit-image: image processing, dedicated to the skimage module.

Tools used in this tutorial:

numpy: basic array manipulation
scipy: scipy.ndimage submodule dedicated to image processing (n-dimensional images). See the documentation:
```
>>> import scipy as sp
```

Common tasks in image processing:

Input/Output, displaying images
Basic manipulations: cropping, flipping, rotating, …
Image filtering: denoising, sharpening
Image segmentation: labeling pixels corresponding to different objects
Classification
Feature extraction
Registration
…

2.6.1. Opening and writing to image files ¶

Writing an array to a file:

importscipyassp
importimageio.v3asiio
f=sp.datasets.face()
iio.imwrite("face.png",f)# uses the Image module (PIL)
importmatplotlib.pyplotasplt
plt.imshow(f)
plt.show()

Creating a NumPy array from an image file:

>>> importimageio.v3asiio
>>> face=sp.datasets.face()
>>> iio.imwrite('face.png',face)# First we need to create the PNG file
>>> face=iio.imread('face.png')
>>> type(face)
<class 'numpy.ndarray'>
>>> face.shape,face.dtype
((768, 1024, 3), dtype('uint8'))

dtype is uint8 for 8-bit images (0-255)

Opening raw files (camera, 3-D images)

>>> face.tofile('face.raw')# Create raw file
>>> face_from_raw=np.fromfile('face.raw',dtype=np.uint8)
>>> face_from_raw.shape
(2359296,)
>>> face_from_raw.shape=(768,1024,3)

Need to know the shape and dtype of the image (how to separate data bytes).

For large data, use np.memmap for memory mapping:

>>> face_memmap=np.memmap('face.raw',dtype=np.uint8,shape=(768,1024,3))

(data are read from the file, and not loaded into memory)

Working on a list of image files

>>> rng=np.random.default_rng(27446968)
>>> foriinrange(10):
... im=rng.integers(0,256,10000,dtype=np.uint8).reshape((100,100))
... iio.imwrite(f'random_{i:02d}.png',im)
>>> fromglobimportglob
>>> filelist=glob('random*.png')
>>> filelist.sort()

2.6.2. Displaying images ¶

Use matplotlib and imshow to display an image inside a matplotlib figure:

>>> f=sp.datasets.face(gray=True)# retrieve a grayscale image
>>> importmatplotlib.pyplotasplt
>>> plt.imshow(f,cmap=plt.cm.gray)
<matplotlib.image.AxesImage object at 0x...>

Increase contrast by setting min and max values:

>>> plt.imshow(f,cmap=plt.cm.gray,vmin=30,vmax=200)
<matplotlib.image.AxesImage object at 0x...>
>>> # Remove axes and ticks
>>> plt.axis('off')
(np.float64(-0.5), np.float64(1023.5), np.float64(767.5), np.float64(-0.5))

Draw contour lines:

>>> plt.contour(f,[50,200])
<matplotlib.contour.QuadContourSet ...>