Skip to content

Model to segment high-center-polygons and troughs using DEM and RGB data.

Notifications You must be signed in to change notification settings

dcruzer231/dem_alaska_model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Semantic Segmentation of DEM data using U-Net architecture

The goal of this project was to segment drone imagery of Alaskan landscape. The landscape was to be segmented into valleys and hills. A U-net model was trained on expert labeled data using dice-loss. In the end the output was not good enough to be used for analysis and the project was halted.

RGB representation of the data

alt text

The DEM data visualized

alt text

The label data

alt text

output using dice-loss

alt text

About

Model to segment high-center-polygons and troughs using DEM and RGB data.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages