KMeans

k-means clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining. k-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, serving as a prototype of the cluster. This results in a partitioning of the data space into Voronoi cells.

Functions

getPalette()

returns a vector of point structs that contain the r,g,b values of the palette.

displayPalette()

displays the original image with its corresponding palette superimposed.

rebuildImage()

after running k-means clustering on the data, we set each point's colour to its centroid's colour.

rebuildGif()

timelapse gif version of rebuildImage()

reduceGif(path, k)

takes a gif and reduce its colour to its k-determined palette.

Dependencies

• CImg 1.6.x

Sample

#include "kimages.h"

int main(){
  kimages foo("\pictures\image.bmp", 5);
  foo.displayPalette();
  return 0;
}

TODO

• Change the palettes of other images to fit new ones.
• Video colour change
• Fix for gifs with matte backgrounds or transparency.