DEVELOPMENT OF NATURAL SOUNDSCAPES – A PATTERN RECOGNITION APPROACH

Authors

  • Ioannis Paraskevas Department of Technology of Informatics and Telecommunications, Technological Education Institute of Kalamata / Branch of Sparta
  • Maria Rangoussi 2Department of Electronics, Technological Education Institute of Piraeus

Keywords:

Soundscape, sound pattern recognition, feature extraction, classification

Abstract

The aim of this work is the mapping of areas of ecological interest focusing on the sound rather than the visual features of this area. Mapping of an area focusing on the visual (morphological) features of the landscape is a mature field with a wide range of applications. On the other hand, ‘soundscapes’ are maps which convey the sound content of an area at a certain time instant. Sound features encapsulate information which can be combined with the visual features of the landscape, in order to derive useful ecological related conclusions. These include monitoring of the wildlife, or the various human activities in the area. Of great interest is the monitoring of the change in the use or inhabitation of the area over time – an aim requiring repeated measurements at regular time intervals. In this work, a method is described for the combination of pattern recognition of the sound recordings with the development of the soundscape of the area of interest

Published

2014-01-29