Publikasjonsdetaljer
- Journal: Proceedings of the Northern Lights Deep Learning Workshop, vol. 3, Monday 28. March 2022
-
Internasjonale standardnumre:
- Elektronisk: 2703-6928
- Lenker:
This paper presents a new method for semi-automatic detection of nature interventions inSentinel-2 satellite images with 10 m spatial res-olution. The Norwegian Environment Agency ismaintaining a map of undisturbed nature in Nor-way. U-Net was used for automated detection ofnew roads, as these are often the cause wheneverthe area of undisturbed nature is reduced. Themethod was able to detect many new roads, butwith some false positives and possibly some falsenegatives (i.e., missing new roads). In conclusion,we have demonstrated that automated detection ofnew roads, for the purpose of updating the mapof undisturbed nature, is possible. We have alsosuggested several improvements of the method toimprove its usefulness.