摘要音乐在人类文明的发展历程中留下了浓墨重彩的一笔，也在今天人们的日常生活中发挥着难以替代的作用，它代表着人们的一种精神追求和享受。从上世纪七十年代起，音乐的特征识别逐渐越来越受到人们的重视，也取得了显著的发展。作为一门交叉学科，它不仅涵盖了计算机科学，信号处理技术和人工智能等等科技学科，也包涵音乐理论，心理学等等人文学科，因此拥有广阔的发展前景和空间。25361
本文选取了西洋乐器中的重要组成部分——钢琴的琴音作为识别样本，通过不同的算法进行乐音的时域和频域分析，并通过matlab这样的工具进行仿真来对比不同算法的性能优劣点。
本文采取的方法主要是基于短时间傅里叶变换的信号分析方法，基于谐波峰值法的信号分析方法与基于小波变换的信号分解与重构方法。在以上三种信号处理方法的基础上，用matlab编写相关程序进行仿真，取用钢琴上88个音键发出的音作为样本分析其时域和基音频域，从而进一步对比算法的性能优劣
关键词 琴音识别 短时傅里叶变换 谐波峰值法 小波变换
毕业论文设计说明书外文摘要
Title Piano sounds Recognition Technology
Abstract
Music in the course of development of human civilization has left an indelible mark, also in today's daily life plays an irreplaceable role, it represents the people of a spiritual pursuit and enjoyment. Since the seventies, feature recognition music gradually more and more people's attention, but also achieved significant development. As an interdisciplinary, it covers not only computer science, signal processing and artificial intelligence, and so on science and technology disciplines, but also encompasses music theory, psychology and so on humanities, it has a bright future and space.
This paper selected Western instruments in an important part - the piano sounds as the identification of samples, time domain and frequency domain analysis tones by different algorithms, and simulation through tools such as matlab to compare the performance of different algorithms strengths and weaknesses .
The method adopted herein is mainly based on short time Fourier transform signal analysis method based on harmonic signal analysis and signal peaks Law decomposition and reconstruction based on wavelet transform. On the basis of these three signal processing methods, using matlab simulation procedures, access to sound on the piano 88 tone key issue as samples to analyze the time-domain and domain-based audio, thus further comparison algorithm performance advantages and disadvantages
Keywords Sounds Recognition STFT Harmonic Peak law Wavelet Transform
目 次
1.绪论 1
1.1 研究的目的与意义 1
1.2 音乐识别技术的国内外发展现状 1
1.2.1 时频分析技术发展 1
1.2.2 单音基音频率识别技术 3
1.2.3 多音基音频率识别技术 3
2. 基本乐理知识 6
2.1 声音的物理本质 6
2.2 基本乐理名词解释 6
2.3 琴音的物理特征 8
2.4 琴音的音乐特征 10
3.乐音信号的时域分析 12
3.1 乐音信号的离散化和预处理 12
3.2 短时能量和短时平均幅度 13
3.3 短时平均过零率 14
3.4 短时自相关函数和短时平均幅度差函数[4] 15
4.乐音信号的频域分析 17
4.1 短时间傅立叶变换 17
4.1.1 短时间傅里叶变换的定义 17 matlab钢琴琴音识别技术研究:http://www.751com.cn/tongxin/lunwen_19104.html