乐音旋律识别研究

Research on musical tune recognition

乐音中的基音成分是进行乐音分析的重要参数,使用小波分解可以比较清晰地反映基音周期情况,FFT对于周期信号有精确的频域分辨.利用小波在信号低频分解方面的良好性质,提出了小波分解的基音检测方法,并用优化的FFT方法获得基音频率的精确定位.在实际开发的系统中取得了较好的效果.

Fundamental frequency is an important parameter in musical notes detection and recognition. This thesis introduces an effective detection method. This method has a good approximate to the musical note signal in low frequency domain and it has a character of noise restraining. This method realizes the precise frequency resolution., It has a high application value.