https://github.com/alexander-lee/k-means-image-compression

Using K-means Clustering for Image Compression
https://github.com/alexander-lee/k-means-image-compression

clustering k-means-clustering k-means-implementation-in-python machine-learning unsupervised-learning

Last synced: 3 months ago
JSON representation

Using K-means Clustering for Image Compression

• Host: GitHub
• URL: https://github.com/alexander-lee/k-means-image-compression
• Owner: alexander-lee
• License: apache-2.0
• Created: 2018-11-08T07:42:59.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-11-08T16:42:05.000Z (over 6 years ago)
• Last Synced: 2025-03-18T11:05:20.875Z (3 months ago)
• Topics: clustering, k-means-clustering, k-means-implementation-in-python, machine-learning, unsupervised-learning
• Language: Python
• Size: 346 KB
• Stars: 9
• Watchers: 1
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Image Compression via K-means Clustering (Color Quantization)

## Overview

This uses the Lloyd's K-means Clustering algorithm with RGB tuples in order to find the `k` colors that best represent the input image. The K-means Clustering algorithm was implemented by me for a better understanding of how the algorithm worked. In addition, this program also produces scatter plots of the colors to better visualize the clusters.

## Running the Program

**NOTE:**  Assuming you are using Python 3.x

```sh

python image_k_means.py  [-k ] [--iterations ] [--save-scatter]

```

* `-k` is the number of clusters you want to find

* `--iterations` is the number of iterations you want to run Lloyd's Algorithm

* `--save-scatter` is a flag to choose to save a Scatter Plot of the data **(WARNING: Very slow)**

## Results

| **Input Image**  | **5-means Clustering (with 20 iterations)** |

| :---:  | :---:  |

|![Sloth](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/Sloth.jpg?raw=true)|![5-means Sloth](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/Sloth_compressed_5.png?raw=true)|

| **10-means Clustering (with 20 iterations**  | **20-means Clustering (with 20 iterations)** |

|![10-means Sloth](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/Sloth_compressed_10.png?raw=true)|![20-means Sloth](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/Sloth_compressed_20.png?raw=true)|

| **Input Scatter Plot by Color**  | **20-means Clustering Scatter Plot by Color** |

| :---:  | :---:  |

| ![Input Scatter Plot](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/Initial_Colors_Sloth.png?raw=true)  | ![20-means Scatter Plot](https://github.com/alexander-lee/k-means-image-compression/blob/master/results/20_Cluster_Colors_Sloth.png?raw=true)  |