Description
By extracting the Mel-frequency cepstral coefficients of the sound signature of a vehicle, it is possible to classify it. The MATLAB neural network algorithm had a high success rate of correctly classifying a vehicle, while the C++ neural network algorithm needed more information.
With the introduction of an additional feature – the physical size of a vehicle – it was possible to achieve a sufficiently high percentage of accuracy in classifying two vehicle classes based on the acoustic characteristics of a vehicle.
This allowed implementing the framework for a reliable classification system using the acoustic and physical features of vehicles.
https://matlab1.com/classification-2/
https://arxiv.org/ftp/arxiv/papers/1804/1804.01212.pdf
Anna –
This project helped me a lot.