Piano accompaniment features and performance processing based on music feature matching algorithm

Piano accompaniment plays an important role in Schubert's art songs. The works render the atmosphere of the works by freehand brushwork, modeling and description, and set off the artistic conception. The perfect music image is created by texture modeling, and it is integrated with melody to achieve a perfect artistic realm. Based on the piano accompaniment features and performance processing of music feature matching algorithm, according to the research and analysis of audio clustering algorithm and the understanding of two-level matching algorithm, this paper puts forward a one-side continuous matching clustering algorithm which automatically groups databases before music matching retrieval. Experiments show that this method can improve retrieval efficiency and ensure retrieval accuracy. Through different tests, the experimental data are analyzed and evaluated, and the performance advantages and retrieval accuracy of the clustering algorithm in this paper are verified.