Research On The Content-Based Musical Retrieval System

The research of content-based musical retrieval is composed of the study of musical theoretics, signal processing and pattern recognition. And it is significant for the management of the music digital library, the music search in the internet and the daily amusement.Some researches about content-based musical retrieval technique are carried out as follows:1. After analyzing the theoretics of content-based musical retrieval, the detailed description and analysis of the new algorithms is shown that is included musical content getting algorithms, musical content description algorithms (musical features extraction algorithms) and features similarity comparison algorithms.2. After studying the musical theoretics, the musical conceptions, the methods to describe the musical features and the theoretics of musical signals analyzing, some signal processing conceptions and algorithms and signal features about music tone extracting is presented.3. An unlimited segmentation algorithm which is according to the vocal segments is presented. The algorithm that is based on the people singing habits segments music film and extracts melody features by using the length of a vocal sentence in the music film as the window length.4. A query by music segment algorithm is presented, that is based on the unlimited segmentation which is according to the vocal segments. The algorithm will compute the tones of single instrumental music segment by pitch estimation, and get the melody features of the music segment. Following the query by music segment algorithm, an exact comparison similarity measure for music segment search is proposed.5. A query by humming algorithm is presented, that is based on the unlimited segmentation which is according to the vocal segments. And the algorithm will get the pitch periods of the humming tones. Following the query by humming algorithm, a obscure comparison similarity measure for humming segment search is proposed.6. According to the algorithms and theories presented in this paper, we designed a model of content-based musical retrieval system, and implemented it. The experimental results show that the system is more exactly in the query by musicsegment and the query by humming.The study work of this thesis provides the preparation for the developing of content-based musical retrieval research and pushes forward the study on melody retrieval.