Feature extraction method of UAV-bird sound signal based on VMD decomposition and M (G) FCC

To solve the problem of difficult classification of UAV and bird acoustic signals, a feature extraction method of noise based on VMD and M(G)FCC is proposed in this paper. Firstly, the ground motion signal of the target is decomposed by VMD to obtain the energy ratio between each IMF signal and the original signal; And then, get GFCC and MFCC from the IMF signal; Finally, the random forest classification algorithm is used to classify the signals, so as to realize the correct recognition of the two.