Seagrass meadows are a very important component of coastal ecosystems, but they are constantly threatened due to anthropic activities. An effective and frequent monitoring of seagrass meadows is therefore widely recognized as an urgent need for their conservation as well as to have an indicator of excessive anthropic pressure on marine ecosystems. This paper explores the use of a deep learning-based image segmentation model for the monitoring of marine seagrass meadows through satellite images. The model can perform a pixel-wise classification of the image, recognizing land, seawater, and seagrass meadows. A dataset of high-resolution satellite images of regions along the Apulian coastline was created and used as training dataset for the deep learning model. The classification performances of the model were then assessed using a set of test images and very promising results were obtained.

Monitoring of Seagrass Meadows Using Satellite Images and U-Net Convolutional Neural Network

Scarpetta, M;Affuso, P;Spadavecchia, M;Andria, G;Giaquinto, N
2022-01-01

Abstract

Seagrass meadows are a very important component of coastal ecosystems, but they are constantly threatened due to anthropic activities. An effective and frequent monitoring of seagrass meadows is therefore widely recognized as an urgent need for their conservation as well as to have an indicator of excessive anthropic pressure on marine ecosystems. This paper explores the use of a deep learning-based image segmentation model for the monitoring of marine seagrass meadows through satellite images. The model can perform a pixel-wise classification of the image, recognizing land, seawater, and seagrass meadows. A dataset of high-resolution satellite images of regions along the Apulian coastline was created and used as training dataset for the deep learning model. The classification performances of the model were then assessed using a set of test images and very promising results were obtained.
2022 IEEE International Instrumentation and Measurement Technology Conference (I2MTC)
978-1-6654-8360-5
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/11589/246620
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • Scopus 2
  • ???jsp.display-item.citation.isi??? 0
social impact