en
Scientific article
Open access
English

Automated Classification of Terrestrial Images: The Contribution to the Remote Sensing of Snow Cover

Published inGeosciences, vol. 9, no. 2, 97
Publication date2019
Abstract

The relation between the fraction of snow cover and the spectral behavior of the surface is a critical issue that must be approached in order to retrieve the snow cover extent from remotely sensed data. Ground-based cameras are an important source of datasets for the preparation of long time series concerning the snow cover. This study investigates the support provided by terrestrial photography for the estimation of a site-specific threshold to discriminate the snow cover. The case study is located in the Italian Alps (Falcade, Italy). The images taken over a ten-year period were analyzed using an automated snow-not-snow detection algorithm based on Spectral Similarity. The performance of the Spectral Similarity approach was initially investigated comparing the results with different supervised methods on a training dataset, and subsequently through automated procedures on the entire dataset. Finally, the integration with satellite snow products explored the opportunity offered by terrestrial photography for calibrating and validating satellite-based data over a decade.

Keywords
  • Fractional snow cover
  • Remote sensing
  • Terrestrial photography
  • Cold regions
Citation (ISO format)
SALZANO, Roberto et al. Automated Classification of Terrestrial Images: The Contribution to the Remote Sensing of Snow Cover. In: Geosciences, 2019, vol. 9, n° 2, p. 97. doi: 10.3390/geosciences9020097
Main files (1)
Article (Published version)
accessLevelPublic
Identifiers
ISSN of the journal2076-3263
411views
171downloads

Technical informations

Creation02/19/2019 12:56:00 PM
First validation02/19/2019 12:56:00 PM
Update time03/15/2023 3:47:11 PM
Status update03/15/2023 3:47:11 PM
Last indexation05/03/2024 7:29:02 AM
All rights reserved by Archive ouverte UNIGE and the University of GenevaunigeBlack