Design And Implementation Of Music Retrieval System Based On Fingerprint Matching

With the development of the computer network,the multimedia resources besides text have been spread freely,especially digital music.Some music workers or music lovers often collect a large number of music material in reserve.Usually the material is cut and mixed into the works,whose original file is not easy to be searched again by text for lack of relevant information.This paper proposes a fingerprint-matching-based music retrieval system,which can retrieve the correct results from the local music library by recording music clips.The music fingerprint is a kind of compact information extracted from the music content,which includes important acoustic characteristics of the music.The system process the time series of the music with Short-time Fourier Transform,in order to calculate a series of "time-frequency" coordinates witch are also the energy peak point,and then make them to pairs as the music fingerprint.In this paper,the key technologies used in the system are studied,including the Fast Fourier Transform,Landmark fingerprinting algorithm,the principle of information retrieval and the fingerprint matching algorithm.On this basis,the system is designed and implemented with Java language.According to the software engineering development flow,the functional requirements determined from the user's point of view,including music recording,the establishment of fingerprint database and music retrieval.Then,the system is divided into several modules to be implemented,such as music recording,fingerprint extraction,fingerprint database and fingerprint matching.We choose two hundred different kinds of mp3 music as a library,each time twenty music are randomly selected to test the system under different conditions like system parameters,recording duration and ambient noise.The retrieval results are all right as long as the recording is longer than ten seconds,which shows that the system has realized the basic function of music retrieval,with good performance in retrieval accuracy,especially in noise immunity.It has certain application prospect and research value.