Análisis de componentes independientes en separación de fuentes de ruido de tráfico en vías interurbanas

Abstract

Traffic noise has turned lately into one of the principal focusses of acoustic pollution, where the identification and classification of noise sources becomes a priority goal. Inside the context of signal processing, nevertheless, the algorithms of blind source separation have been applied traditionally to voice and music signals. In this context of application, the role played by the noise signals is identified always by not wished sources, intending to suppress their contribution from the the acquired signal. From a different point of view, respect to the treatment of the noise signals, this article approaches the different strategies followed in the application of "Independent Components Analysis" techniques, orientated to traffic noise sources blind separation. The study carried out reflects interesting results respects to the quality in the separation and later classification processes.