NOTE: Image quantized to 5 discrete colors

Color quantization is quantization applied to vectors in a three dimensional color space. In this case the color palette choosen for color quantization, is obteined through performing the k-means algorithm with the RGB values of each pixel in an image, as points in a three dimensional space. Afterwards the values of each pixel are rounded or quantized to the nearest centroid obtained from the k-means algorithm.

This algorithm can be applied to a bitmap image, and produces an output bitmap image with n different colors, trying to conserve the original colors as acurately as possible.

1. Build with

   make

2. Run the program with

   ./main <src image> <out image> <num colors>