Particle Filter Animal Voice Recognition Method Based on the Spectrum Modeling and Synthesis

  • Haifeng Duan
Keywords: Spectrum Modeling, Particle Filter, Animal Voice Recognition, Signal Processing


With the indiscriminate fishing, deforestation and pollution in recent years, the number of wild animals
has been greatly reduced, and the call for the protection of wild animals is also increasing. But in daily life,
people can only hear the calls of animals and do not see them, which makes it difficult to protect species. Therefore,
the research of species recognition technology based on animal call is of practical significance for the protection
and determination of local species, as well as for biological research and environmental monitoring.
Therefore, species recognition based on animal call is of practical significance. Species recognition technology
based on animal call will become a hot issue in the field of biometric recognition. In this paper, based on the image
processing method, a two-dimensional array feature mapping input speech signal is constructed, and the particle
filter is used for speech recognition. In order to improve the effect of speech recognition, a learning method
of local finite weight shared particle filter is established according to the characteristics of speech signal. In this
paper, through the classification and training of pig calls under different behaviors, the algorithm has high
recognition efficiency, and can accurately recognize the basic pig calls, which improves the recognition rate. It
is proved that the algorithm in this paper is also feasible in the recognition of other animal speech, and animal
speech recognition provides an effective guarantee for animal behavior analysis and animal protection.