Different techniques can be applied for shoreline acquisition. Direct survey, based on GNSS (Global Navigation Satellite System) or total station, permits to obtain 3D information that is useful for the correct definition of the coastline also in consideration of the tidal effects. However, the acquisition of long stretches of coast using in-situ survey may be too expensive and time consuming. Additionally, many studies require to reconstruct temporal shoreline dynamics, and, in absence of survey carried out in the past, remotely sensed data may be a valuable source of information. For those reasons, there is a widespread usage of aerial and satellite imagery in many studies needing coastline detection. This research aims to analyze methodological aspects of coastline extraction from optical satellite imagery at medium and high resolution: the evaluation of the results accuracy permits to compare two different approaches based on the multispectral band use. The attention is focused on Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI), both applied to medium resolution imagery (Landsat 8 OLI) and to high resolution imagery (GeoEye-1). Maximum Likelihood Classification (MLC), one of the most common classification methods in remote sensing based on Bayes’ Theorem, is applied to determine a threshold to separate seawater from land. An index based on the direct comparison between the automatic extracted coastline and the manually delineation of it, is used to evaluate the accuracy of the results. Both indices permit to obtain acceptable results reporting accuracy values less than the pixel dimension. However, the accuracy level of NDWI is slightly higher than NDVI.

Coastline Extraction from Optical Satellite Imagery and Accuracy Evaluation

Alcaras E.;Errico A.;Falchi U.;Parente C.;Vallario A.
2020-01-01

Abstract

Different techniques can be applied for shoreline acquisition. Direct survey, based on GNSS (Global Navigation Satellite System) or total station, permits to obtain 3D information that is useful for the correct definition of the coastline also in consideration of the tidal effects. However, the acquisition of long stretches of coast using in-situ survey may be too expensive and time consuming. Additionally, many studies require to reconstruct temporal shoreline dynamics, and, in absence of survey carried out in the past, remotely sensed data may be a valuable source of information. For those reasons, there is a widespread usage of aerial and satellite imagery in many studies needing coastline detection. This research aims to analyze methodological aspects of coastline extraction from optical satellite imagery at medium and high resolution: the evaluation of the results accuracy permits to compare two different approaches based on the multispectral band use. The attention is focused on Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI), both applied to medium resolution imagery (Landsat 8 OLI) and to high resolution imagery (GeoEye-1). Maximum Likelihood Classification (MLC), one of the most common classification methods in remote sensing based on Bayes’ Theorem, is applied to determine a threshold to separate seawater from land. An index based on the direct comparison between the automatic extracted coastline and the manually delineation of it, is used to evaluate the accuracy of the results. Both indices permit to obtain acceptable results reporting accuracy values less than the pixel dimension. However, the accuracy level of NDWI is slightly higher than NDVI.
2020
978-3-030-62799-7
978-3-030-62800-0
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11367/88781
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 12
  • ???jsp.display-item.citation.isi??? ND
social impact