Jamie Walker - Python

This repository is just a collection of my python code that I've developed. Feel free to use any of it if it may be helpful for you...

KMeansClusterGray.py

This is an implementation of the K-Means clustering algorithm I made for grayscale images. It was initially made for grayscale images with backgrounds of all 0 values. It ignores the background and therefore it actually makes k+1 clusters.

Documentation / How to use

Create a new instance of the KMeansClusterGray class, passing in the input filename and the amount of clusters to the constructor.

clustering = KMeansClusterGray(filename, 3)

Call the KMeansClusterGray.convert(n_max_iterations) function.

The function takes in the maximum number of iterations to perform. It returns a numpy array of the clustered image and a numpy array containing the mean values for the clusters.

clustered_image, cluster_values = clustering.convert(200)