复杂噪声场景下的电子音乐信号智能检测算法设计  

Design of intelligent detection algorithm for electronic music signal in complex noise scene

作　　者：许晶晶[1] XU Jingjing(Chongqing Nanfang Translators College of SISU,Chongqing 401120,China)

机构地区：[1]四川外国语大学重庆南方翻译学院,重庆401120

出　　处：《现代电子技术》2020年第19期49-52,57,共5页Modern Electronics Technique

摘　　要：电子音乐信号检测有利于音乐的分类和后续处理,针对当前电子音乐信号检测方法没有考虑噪声对电子音乐信号检测结果的干扰,导致电子音乐信号检测错误率大,检测时间长,无法满足现代大规模电子音乐信号检测的实际要求。为了解决当前电子音乐信号检测过程中存在的一些局限性,提出一种复杂噪声场景下的电子音乐信号智能检测算法。首先分析电子音乐信号智能检测的原理,引入小波变换对电子音乐信号中的噪声进行抑制;然后从消噪的电子音乐信号中提取特征,并根据电子音乐信号类型构建学习样本;最后引入最小二乘支持向量机对电子音乐信号智能检测模型进行构建,分析电子音乐信号的变化特点,并对其参数进行相应的优化。仿真实验结果表明,对于复杂噪声场景下的电子音乐信号,文中方法可以获得较高精度的检测结果,检测时间也明显得到了减少,获得了比其他方法更优的电子音乐信号,对复杂噪声具有较强的鲁棒性。Electronic music signal detection is conducive to music classification and follow-up processing.For the current electronic music signal detection method,the interference of noise on the electronic music signal detection results is not considered,which makes the electronic music signal detection error rate large and detection time long,and is unable to meet the actual requirements of large-scale modern electronic music signal detection.In order to solve some limitations existing in the process of the current electronic music signal detection,an intelligent detection algorithm of electronic music signal in complex noise scene is proposed.The principle of intelligent detection of electronic music signal is analyzed.The wavelet transform is introduced to suppress the noise in electronic music signal,and then the features of the denoised electronic music signal are extracted.The learning samples are constructed according to the types of electronic music signal.The least square support vector machine is introduced to construct the intelligent detection model of electronic music signal,analyze the changes of electronic music signal,and carried out corresponding optimization of the parameters.The simulation results show that,for the electronic music signal in the complex noise scene,the method proposed in this paper can obtain high-precision detection results,and its detection time is significantly reduced,and its electronic music signals are better than those of other methods,which has strong robustness to complex noise.

关 键 词：电子音乐信号检测 智能检测算法 检测原理分析 噪声抑制 模型构建 参数优化 

分 类 号：TN911.23-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]