Welcome to pycimg’s documentation!¶

README¶

pycimg is a python extension for the CImg library.

The package contains a single class CImg that provides access to the image processing methods of the CImg library.

Pixel data of CImg objects can be accessed as a numpy array.

Vice versa new CImg objects can be created from pixel data in a numpy array or an image file. Supported file formats are png, jpeg, tiff, bmp, and cimg.

from pycimg import CImg
import numpy as np

# Load image from file
img = CImg('test/test.png')
img.display()

# Access pixel data as numpy array
arr = img.asarray()
# Set pixels in upper left 100 x 100 px rectangle
arr[:,:,0:99,0:99] = 0
# Pixel data is shared with the image instance
img.display()

# Create image from numpy array
img = CImg(np.random.randn(100,100))

Features¶

Access pixel data as a numpy array.
Builtin support for reading/writing png, jpeg, and tiff image formats.

Installation¶

Install pycimg by running:

pip install pycimg

Documentation¶

See readthedocs.

License¶

The project is licensed under the GPL3 license.

API¶

class pycimg.pycimg.CImg(*args, **kwargs)[source]¶

CImg is a wrapper class for the CImg library:

__eq__(img)[source]¶: Return self==value.

__init__(*args, **kwargs)[source]¶

Create CImg with given data type.

Supported datatypes are int8, int16, int32, uint8, uint16, uint32, float32, and float64.

Examples

1. Create empty image with default type float32 im = CImg()

2. Create image from file. im = CImg(“filename.png”)

3. Create image from numpy array arr = np.zeros((100,2)) im = CImg(arr)

4. Create image of size 100x200 with data type uint8 im = CImg((100, 200), dtype=uint8)

Keyword Arguments:
Parameters:	image filename, numpy array, or image size. (Either) –
	dtype – Data type of CImg.
Raises:	`RuntimeError` – For unsupported data types.

__repr__()[source]¶: Return repr(self).

__str__()[source]¶: Return str(self).

__weakref__¶: list of weak references to the object (if defined)

asarray(copy=False)[source]¶

Returns image data as a numpy array.

Parameters:	copy (bool) – False.

depth¶: Return depth of image.

height¶: Return height of image.

isempty()[source]¶: Return true if image is empty.

shape¶: Return shape of image data.

size¶: Return the total number of pixel values in the image.

spectrum¶: Return spectrum (number of channels) of image.

width¶: Return width of image.