TerrainSR: Automated High‑Resolution Terrain Generation from Low‑Resolution Elevation Data
Problem
Developers need to convert low‑resolution historical elevation data (e.g., 100 m DEM) into detailed, natural‑looking high‑resolution terrain for a 1:1 historical game world; they also want to generate multiple map variants from the same low‑resolution seed to preserve gameplay balance and ensure uniqueness, but current solutions struggle with massive storage, fragmented high‑resolution data, and artifacts from man‑made objects obscuring natural terrain.
Pain points
- Low‑resolution terrain data lacks sufficient detail
- Storing and processing massive, fragmented high‑resolution data is difficult
- Man‑made objects (buildings, mines, etc.) obscure natural terrain, complicating reconstruction
- No ability to generate diverse map variants from the same low‑resolution seed to meet gameplay balance requirements
Questions to verify
- Does it support generating realistic terrain for non‑Earth planetary bodies such as Mars?
- Would users be willing to pay or subscribe for generating diverse map variants?
- How does the current model perform in terms of processing speed and storage requirements on large real‑world terrain datasets?
Evidence
quote · @Anonymous
Really cool! I did my master's thesis on this same problem, upscaling low resolution DEM, though my approach used a high resolution color image to guide the process.
quote · @Anonymous
On my phone but very interested in this (hence leaving a comment so I can find it later). What’s the variation, can we generate different maps from the same low res seed?I’m interested in this because “macro maps” can be hand built in a way that may want to preserve gameplay balance while individual games can still feel broadly unique.
quote · @Anonymous
can you talk to some about how you trained this? this is such a neat idea!!
fact
100m data was too low-res while 10m LIDAR data was patchy, took hundreds of GBs to store and was full of manmade objects like mines, buildings and so on.
Disclaimer: this card is distilled by AI from community discussions as “pain point evidence”. It is a starting point for need analysis and validation, not a “validated market need”. AI has analytical limits; please open the original discussion above to manually verify before key decisions.