Aircraft take-off noises classification based on human auditory's matched features extraction

Air transportation is one of the most important services in the world, contributing greatly to the advancement of modern society. However, it has a local and a global impact on the environment making aircraft take-off noise an important environmental public health concern near airports, and this is a significant subject for monitoring and research. In this work an experimentally validated computational model for aircraft classification is presented. In addition, potentially harmful effects to human health and comfort associated with noise exposure are discussed. The feature extraction of aircraft take-off signals is conducted through a 1/24 octave analysis and Mel frequency cepstral coefficients (MFCC). The aircraft classification is made by using two parallel feed forward neural networks. The aircraft are clustered into classes depending on the installed engine type. This model has 13 aircraft classes and a classification level above 83% with measurements in real time environment.