How can we best handle the results from AI segmentation from different data sources

Abstract

The thesis is focusing on change detection of building footprints and how DL methods can be implemented in order to make this process more effi cient, at the same time as looking at the credibility of such implementation, and the expected performance with further improvements and use on bigger applications in the future.

The thesis is focusing on change detection of building footprints and how DL methods can be implemented in order to make this process more effi cient, at the same time as looking at the credibility of such implementation, and the expected performance with further improvements and use on bigger applications in the future.