Research On Feature Recognition And Representation Technology Of Music Signal Based On DSP

The digitalization and analysis processing technology of music signals is the core of digital music technology.This paper studies the feature recognition and reproduction technology of music signal based on DSP,aiming at designing a system to help music learners to learn music and create music.This paper studies the modeling and simulation of music signals and their analysis and processing,and the design of embedded systems for the actual needs of music learning and creating,and lays a foundation for the study of automatic arrangement and automatic notation.Firstly,this paper studies the modeling and simulation of music signal and its analysis and processing algorithm.Taking the piano as the research material and combining the four elements of music sound,it analyzes and extracts the characteristic parameters of piano notes and establishes the mathematical model of single note signal and music score signal.The single note recognition algorithm is studied,the Mel frequency cepstrum coefficient of the signal is extracted,and the DTW algorithm is improved to realize single note recognition.On the basis of the realization of the single-note algorithm,combined with the note-time value segmentation method based on the energy-entropy-ratio,the score is segmented into single-note sequences to realize the score recognition.Next,this paper studies the music signal reproduction algorithm and performs simulation.The music signal generation algorithm is combined with the music score to perform simulation to generate a music sequence signal,and then converted into a playable data format to realize music signal reproduction of single notes and continuous notes.Based on the completion of music signal modeling and algorithm research,the hardware and software design of DSP device system based on BF609 is given.This paper first designs the overall hardware structure of the system,and designs the circuit according to the functional requirements of the system,and gives detailed design for the main modules of the system hardware,namely audio acquisition module,audio playback module,storage module and human-computer interaction module,in which the audio codec SSM2603 is used to integrate the audio acquisition module with the playback module.The system software design is based on the hardware platform.The algorithm implementation flow of each functional module of the system is given,including preprocessing algorithm,feature parameter extraction algorithm,recognition algorithm and music synthesis algorithm,which algorithm programs are ported from the aforementioned algorithm emulator.Finally,the DSP system is tested and analyzed,including module testing,system overall functional testing and recognition rate testing.Among them,the module test includes audio line output test,audio acquisition module test,and recognition algorithm test.On the basis of correct module test,the whole system is functionally tested to verify the system working status The recognition rate of the system is tested with several sets of known score data,and the recognition result is compared with the correct result to judge the recognition performance of the system.The test data shows that the DSP embedded system designed by this paper has a recognition rate of 93% for continuous notes.The test of the music synthesis algorithm is based on multi-person evaluation,and it is confirmed that the algorithm has achieved good piano synthesis effect.