Resumen
System and procedure for detecting and identifying sounds in real time produced by specific sound sources, comprising: sampling, an analog audio signal captured by a sensor means (1) from at least one objective sound source (2); digitize the sampled signal; digitize the digitized signal to obtain frames of a stationary character; extract, for each frame, at least one feature vector; classify extracted feature vectors; standardize the output of the classification; store the normalized output signal; soften the stored normalized output signal and mark and extract sound events of the smoothed signal from parameters configured by the user. Additionally, it comprises an optimization stage that allows obtaining statistical indicators in the detection and joining temporally close sound events according to a parameter configurable by the user. (Machine-translation by Google Translate, not legally binding)