| The origin of this paper is the practical demands of the professional songsmiths. With the development of speech recognition and computer technology, they hope to apply computer technology to recognize the real-time musical sound and then finish memorizing automatically. How to realize the recognition of the musical notes is the main research aspect in this paper.The technology to let computer understand the sound being played on the musical instrument has important values in composing at the computer. However, the existed technologies only localize in the musical sound sampling, signal filtering, sound effects processing and the musical sound editing. The existed system uses MIDI keyboard to realize musical notes input and does not deal with the note recognition discussed in this paper. There are no concerned researches which aimed at the musical note recognition have been developed nowadays. Through the beforetime market research among the musical professional people, a potential vendibility is found. Combining with these practical demands, and beginning from the researches of the physical and musical characters of the musical sound, a recognization method is advanced; it is an effective and available in practice.The musical note recognition method presented in this paper considers the steady time domain periods trait of a signal note, and the strongpoint of DWT which decomposes the signal to low bandwidth rapidly. The low-pass component of DWT is approximate to the original signal, and the period of the low frequency remains is hold perfectly, thereafter, uses FFT to legalize the precise frequency resolution and acquires perfect recognition results. This method can be redounding to the musical database searching.There is little mature technique about the musical note recognition; therefore, this paper tries to be a path breaker. This paper has built a whole piano musical notes library, and finished the related theory study and the design of an application system.
|
PDF Full Text Request