Acoustic vehicle classification based on the optimum packet energy and fuzzy classifier

This paper proposes a method of acoustic vehicle classification. We extract acoustic feature of wheeled vehicle and track vehicle based on the optimum wavelet packet energy and classify these two types of vehicle based on the fuzzy classifier. For evaluation purposes, real data are used from DARPA's SensIT project. The experiment results show that the proposed method can improve the correct recognition rate, and achieve the balance between the processing time and performance.