Seabed classification using supervised functional data analysys techniques

Resumo

The objective of this work is the numerical analysis of the discretization parameters used in the functional statistical methodologies, on which the supervised classification for the automatic identification of seabed types in coastal zones is based. This methodology uses acoustic data obtained by a simple beam echo sounder (at 38kHz) coupled to a small boat. Each of the acoustic intensity curves has been previously preprocessed by applying time, power and ping length corrections in order to eliminate its dependence on depth. The experimental data were obtained in a controlled environment in the region of Cabo de Palos (Murcia, Spain), studying three different types of bottom: sandy, sandy with vegetation and rock. The statistical techniques applied to this particular case belong to the group of supervised classification techniques but combined with functional data procedures. The numerical results obtained and its analysis confirm that the use of a low number of elements of the discrete basis combined with their accurate approximation properties provide a correct classification of the three types of seabed considered.