AI-Powered Interactive Visualization for Landscape Evolution in Virtual Reality
Published in International Journal of Human-Computer Studies (Under Review), 2025
Historical maps encode valuable information about landscape evolution, yet they require critical map reading and spatial thinking skills to decode and interpret this information, which the general public often lacks. We introduce an AI-powered visualization system that automatically transforms a series of historical maps into an immersive animation of changing landscapes in VR, offering a new way to re-imagine static archival data as dynamic, interactive experiences. This system, leveraging computer vision techniques and generative AI, enables scalable and customized reconstruction of 3D landscape models from 2D historical maps. To further support spatiotemporal understanding, our system incorporates interactive features like temporal interpolation to create evolution animation, terrain exaggeration, and interactive controls for scale and lighting conditions. The system’s effectiveness is evaluated through a user study and expert interviews, demonstrating its usability for terrain understanding and landscape feature recall. The findings highlight the potential of our system as a communication and education tool in geography and spatial planning.
Recommended citation: Mischa Bauckhage, Sidi Wu*, Tianyi Xiao*, Jiakun Xu, Martin Raubal, Lorenz Hurni, Yizi Chen. (2025). AI-Powered Interactive Visualization for Landscape Evolution in Virtual Reality. International Journal of Human-Computer Studies (Under Review).
Download Paper