Converting map images from real-world data to match fictional vector data VIA artificial intelligence and image processing techniques

The U.S. military and its allies have a need for fictional terrain paper map images suitable for use during simulated training exercises. They currently have the capability of rapidly creating fictional vector data for modeling and simulation using real world source data. However, the end product from this capability is not suitable for use as the basis of a fictional paper map image. If the fictional vector data used by the simulations to stand in for the real world were used to create a paper map, it would not reflect the errors that exist between the real world and commercial paper maps. This thesis will explore different approaches to achieving the goal of creating fictional map images from real world map images using combined artificial intelligence and image processing techniques while attempting to maintain the error in the maps and the fog of war.