Recognition of piano keyboards based on sound feature extraction

Abstract

School of Computing, University of Kent, CT2 7NZ, England, UK (Received May 10 2019, accepted June 11 2019) The tones of piano are produced by regular periodic oscillations and different keys\u2019 individual inherent \u00a0audio \u00a0characteristics. \u00a0Based \u00a0on \u00a0the \u00a0collection \u00a0of \u00a0the \u00a0sound \u00a0wave \u00a0signals \u00a0and \u00a0the \u00a0analysis \u00a0of \u00a0the frequency characteristics from different keys, this study explores the methods to effectively identify and detect the corresponding numbers of keys. This paper will also compare and analyzed the different methods of audio signal extraction and identification of piano keys, to find effective methods to make theoretical preparation for further system development.