This is a python package to compress images in python using KMeans-Clustering technique.

Download the package using:

pip3 install KMeansImage

from KMeansImage import kmeans_image
k_img = kmeans_image( $image_path_with_image_name ,  $output_colors )
k_img.save( $save_path_with_file_name )

In case you are using Jupyter notebook, you can also use:

from KMeansImage import kmeans_image_ui
k_img = kmeans_image_ui( $image_directory_path )
k_img.save( $save_path_with_file_name )

Import Library

from KMeansImage import kmeans_image

First parameter = path
Second parameter = No. of colors i.e. Clusters

k_img = kmeans_image('/home/sahil/Downloads/KMeansImage/res/batman.jpg', 20)

Original Image size: 419.291KB
Compressed Image: 253.235KB

See the difference between size of both pics
For saving the file use -

k_img.save()

Image saved: /home/sahil/Downloads/KMeansImage/res/batman_KMeans_.jpg

from KMeansImage import kmeans_image_ui

Parameter = Image directory path

k_img = kmeans_image_ui('/home/sahil/Downloads/KMeansImage/res/')

k_img.save()

Image saved: /home/sahil/Downloads/KMeansImage/res/sherlock_KMeans_.jpg

pip install ipywidgets
jupyter nbextension enable --py widgetsnbextension
jupyter labextension install @jupyter-widgets/jupyterlab-manager

You can also download using Github:

• Download and extract this repo
• Move the terminal to the directory of setup.py

pip install .

• numpy
• scikit-learn
• matplotlib
• ipywidgets

Sahil

• Add a button for saving images instead of save() in interactive notebook
• Add some boundary conditions
• Add Test cases

MIT

Free Software, Hell Yeah!