Improvements to Ambient Audio Collection Systems for Smart City Applications
Internet of Things (IoT) has introduced the ability to integrate sensors into everyday life. Ambient audio collection systems are an example of useful IoT and allow citizens to visualize their environment based on auditory parameters. This can be utilized to positively affect areas such as real estate, city planning, and daily activities. There are many challenges that need to be overcome to have a system that meets the power and accuracy requirements. In this thesis, an ambient audio acquisition system was designed and improved using a custom threshold circuit and filtering. The results show a power reduction of up to 70% when compared to a standard implementation. The accuracy was validated by using magnitude-squared coherence. The Boll, Berouti, and Kamath spectral subtraction algorithms were applied to urban sound audio clips. This was shown to increase the accuracy and improve the Signal-to-Noise Ratio (SNR) by up to 20 dB when compared to an unfiltered signal. This also showed an urban sound classifier accuracy improvement of 10% in high-noise environments.