Advertisemen
Decided to see how well my clustering implementation would work for landscape generation (for my coursework), so into Java I ported it.
The application works by clustering all the obstacles to a pre-defined amount of clusters (currently 12), and then checking which cluster each square of the map is closest to a generating a different landscape for each cluster. Since there are more clusters than landscape types, the cluster indexes are modulo the landscape type number.
Below is the landscape generated:
Advertisemen
Tidak ada komentar:
Posting Komentar