Research On Detection Technology Of Piano Music Signals

With the improvement of people’s living standards,more and more people study an instrument to enhance the temperament accomplishment.However,beginners need to spend lots of time in the study of piano correction.In this paper,the detection technology of piano music signal is researched,aiming at analyzing the characteristics of musical signal of piano playing,combining with the information of prior musical score,detecting and recognizing the music signal.The main work is as follows:1)The application of time-frequency analysis technique in piano music signal is studied.The time-frequency spectra of music signals based on STFT,CWT and HHT are analyzed and simulated.Wavelet decomposition and EMD decomposition are discussed,which laid a foundation for the research of music detection technology.2)Discussed the onset of the note detection.Based on the characteristics of short-time energy,high-frequency content(HFC),frequency variance and spectral difference of music signal,the onset detection based on different signal characteristics was analyzed with single tone music and multi-tone music simulation.The simulation results show that HFC is more suitable for the onset detection of piano music.3)The fundamental frequency detection method of tone music is discussed.A detection method based on spectral peak ordering is proposed.The simulation shows that the method can detect the fundamental frequency of the whole frequency band.4)Based on the idea of signal decomposition,this paper proposes a method of multi-frequency detection based on subband decomposition and FFT.First,the spectrum peak sorting method is extended to the multi-frequency detection field,and the two-channel processing is realized.Secondly,the detection method of binary wavelet transform and FFT is discussed.Then,the EMD and FFT are combined to detect according to the characteristics of EMD adaptive decomposition.The simulation results show that EMD and FFT can achieve accurate detection of multiple fundamental frequencies.At last,the detection and recognition of the whole music are completed with the score.The test results show that this method has a better detection effect on piano music.