CESIMAR - CENPAT   25625
CENTRO PARA EL ESTUDIO DE SISTEMAS MARINOS
Unidad Ejecutora - UE
artículos
Título:
Comparative Study of FDA and Time Series Approaches for Seabed Classification from Acoustic Curves
Autor/es:
TARRÍO-SAAVEDRA, JAVIER; PRIETO, ANDRÉS; SÁNCHEZ-CARNERO, NOELA
Revista:
MATHEMATICAL GEOSCIENCES
Editorial:
SPRINGER HEIDELBERG
Referencias:
Año: 2019
ISSN:
1874-8961
Resumen:
Seabed classification in coastal environments is usually accomplished using multivariate methods applied to acoustic features from corrected or uncorrected echoes. This paper presents a comparative study of alternative statistical tools based on time series clustering and non-hierarchical clustering methods for functional data. This allows us to consider the entire acoustic signal without information reduction and assess performance using data acquired in a controlled environment with three different seabed types. The methods considered are used to both analyse the classification power of the recorded echoes and identify the most significant portions of signal.