Datascience for NMCAS
This virtual event aimed to raise awareness and understanding of datascience within the context of NMCAs. By sharing datascience usecases, technical challenges in datascience projects and experiences of implementing datascience projects to production, this two-day workshop helped us to appreciate the impact, benefits and effect of datascience on the business processes of NMCAs.
All datascience projects in this event focused on geospatial data, for example: remote sensing, satellite and aerial imagery and LIDAR.
The workshop was aimed at staff members of NMCAs and researchers at NMCAs and academic institutions with an interest in datascience in the geospatial sector.
This free virtual workshop was spread over two half-day sessions on 2 & 3 June 2022.
Recordings: You can watch the recordings of the workshop here.
Programme
Day 1: Introduction to datascience & presentation of usecases
Day 1: 2 June 9:30 – 13:00 (CET) | ||
09:30 | Welcome from EuroSDR | Martijn Rijsdijk |
09:45 | Introduction to datascience in the context for NMCAs Kadaster, Introduction to datascience for NMCAs |
|
10:15 | Usecases of data datascience research - IGN, Creating national semantic segmentation datasets for land cover - University of Copenhagen, Tree detection with remote Sensing and Neural networks |
|
11:05 | Break |
|
11:25 | Usecases of data datascience projects - Styrelsen for Dataforsyning og Effektivisering (SDFE), How can Machine Learning and Free Open Geospatial Data benefit each other? - GUGiK, Deep learning for automatic detection of errors in aerial imagery - National Land Survey of Finland, Building detection using the UNet and the results of expert evaluation |
Emilia Hattula |
12:40 | Wrap up and closure Closing of the day; retrospect and lessons learned |
|
Day 2: Technical challenges & implementing datascience projects to production
Day 2: 3 June 9:30 – 13:00 (CET) | ||
09:30 | Welcome Day 2 | Martijn Rijsdijk |
09:35 | Technical challanges in datascience projects - Kadaster, Reconstructing the Dutch Cadastral Map - National Land Survey of Finland, Problems and challenges of deep learning in the ATMU project - Kartverket, How to reproduce colors in-between projects to facilitate machine learning color analysis? |
|
10:50 | Break |
|
11:10 | Implementing datascience projects to production - Fondazione Bruno Kessler, 3D powerline classification for vegetation occlusion detection - OPEGIEKA, From proof of concept to real-time application challenges in in implementation of a novel AI based solution for Lidar point cloud classification - Fondazione Bruno Kessler, Upscaling photovoltaic potential on the entire Trentino region |
|
12:40 | Wrap up and closure Closing of the day; retrospect and lessons learned |
|
Organising commitee
The following persons were involved in the preparation of this workshop:
- Anna Bober, Head Office of Geodesy and Cartography (GUGIK), Poland
- Martijn Haanappel, Kadaster, The Netherlands
- Martijn Rijsdijk, Kadaster, The Netherlands
- Matthieu Porte, IGN, France
- Linghli Zhu, National Land Survey of Finland (Maanmittauslaitos), Finland