How to use combinatorial optimization problems (Traveling Salesman Problem) for procedural landscape generation

In this paper, we examine how the traveling salesman problem (a combinatorial optimization problem) can be used to create virtual landscapes. For this work, we show how the entire search space of ten city TSP instances can be organized into virtual landscapes, and illustrate how by changing the TSP instance problem we can control some properties of these landscapes. We provide three different methodologies for producing landscapes and show results for TSP problem instances. Our results show that our one of our methodologies generates the best aesthetically looking results and can be controlled by the problem.