Identification of Ambulance Siren sound and Analysis of the signal using statistical method

Identification of ambulance sound in a busy traffic is a challenging task. The Ambulance siren sound should be segregated from noisy signal in frequency domain. In this paper, the sound signal is first captured by the onboard mic which is then processed in Python by the Sound device module. The identified ambulance signal is compared with Horn, Wail, Yelp, and Hi-Lo ambulance siren. These signals are analyzed before performing any further statistical analysis. Sampling frequency of 10KHZ is chosen to investigate the data in the bandwidth of 500-1500Hz. The data is then formatted, after which the statistical model is used to find Mean and Standard Deviation.