Image Compression

• Given a RGB bitmap image file, clustered the pixels using clustering algorithms
• Created the final compressed image by coloring each cluster with the representative color of it’s centroid.

Clustering Algoithms Used

• K-means
• K-medoids

K-Medoids

• Manhattan Distance Used as Distance Metric
• Partitioning Around Medoids (PAM) Algorithm

How to run

• run mainprogram(filename, no of clusters)
• Eg: run mainprogram(‘beach.bmp’, 3)