A high frequency approach for seabed vegetation characterization

Abstract

This paper focuses on the characterization of seabed vegetation, an interesting problem in a lot of Mediterranean sea coastal sites since water quality and pollution degree can be efficaciously solved by investigating the sea flora extents. The acquisition of the data about marine vegetation is accomplished by using a high frequency imaging sonar scanning the seabed. The proposed approach is based on a statistical feature-based description of the raw sonar data. A simple classifier attempts fo characterize the acquired samples.