Item: FROM NOTEBOOK TO CAAML — DIGITIZING HANDWRITTEN SNOW PROFILES
-
-
Title: FROM NOTEBOOK TO CAAML — DIGITIZING HANDWRITTEN SNOW PROFILES
Proceedings: International Snow Science Workshop 2024, Tromsø, Norway
Authors:
- Aron Widforss [ Norwegian Water Resources and Energy Directorate ]
- Øystein Myhre [ Norwegian Water Resources and Energy Directorate ]
- Erik Edin [ Norwegian Water Resources and Energy Directorate ]
Date: 2024-09-23
Abstract: Most digital snow profile solutions have an unergonomic data entry step, and manually converting profiles from notebooks to a digital representation is double work. The Varsom Snow Profile Scanner uses machine learning powered object and character recognition to parse the notebook and deliver a digital profile in standard CAAML format. Using standardized snow observation notebooks from Avalanche Warning Service Tirol we show that this technology can be used in a workflow to digitize snow profiles, even when neither the notebook nor notes were created with machine interpretation in mind. However, such a workflow would need a human in control to fix any mistakes of the machine. If integrated in an operational snow observation app (e.g. https://regobs.no/), this technology will change the observer’s role in the digitization process from data entry to quality control. By minimizing the amount of time spent on it we hope to make data entry a less strenuous task and give observers more time in the snow.
Object ID: ISSW2024_P1.25.pdf
Language of Article: English
Presenter(s): Aron Widforss
Keywords: CAAML, snow profile, digitization, snow observations, machine learning
Page Number(s): 225 - 232
-