蛋鸡声音信号去噪方法对比分析
发布时间：2018-08-06 09:48
【摘要】：声音分析技术已成为研究动物行为、动物福利的一种重要工具,蛋鸡声音可用来评价其健康或福利状况,然而规模化蛋鸡舍中存在较多风机噪声等声源干扰,这对准确识别蛋鸡发声有很大影响。本文以海兰褐蛋鸡为例,预处理获取的声音信号,以减少风机噪声的干扰。利用数字化声音采集平台采集不同类型的蛋鸡发声和风机噪声音频,采用Lab VIEW软件进行声音信号处理并分析蛋鸡声音和风机噪声的时频特征。同时,对比分析不同去噪方法(IIR滤波器去噪、小波阈值去噪和改进谱减法去噪)在去除风机噪声方面的效果。结果表明,在信噪比为-8~20 d B声音环境下,改进谱减法均方根误差最小(0.03~0.38),算法运行耗时最短(6~7 ms),在实际应用中去噪效果较好,可为规模化蛋鸡舍中风机噪声环境下的蛋鸡声音信号处理和分析提供参考。
[Abstract]:Sound analysis technology has become an important tool for the study of animal behavior and animal welfare. The sound of laying hen can be used to evaluate its health or welfare status. However, there are many noise sources such as fan noise in the large chicken house, which has a great influence on the accurate identification of the sound of laying hens. The sound signal is used to reduce the noise of the fan. Using the digital sound acquisition platform to collect the sound of different types of laying hens and the noise and audio of the fan, Lab VIEW software is used to carry out the sound signal processing and analyze the time frequency characteristics of the sound and fan noise. At the same time, the different denoising methods (IIR filter denoising, wavelet threshold value) are compared and analyzed. Noise and improved spectral subtraction denoising effect in removing fan noise. The results show that, under the noise ratio of -8~20 D B, the improved spectral subtraction mean square root error is minimal (0.03~0.38), the algorithm running time is shortest (6~7 MS), and the effect of noise removal is better in practical application, and it can be the sound of the egg noise under the noise environment of the large-scale egg chicken house. Signal processing and analysis provide reference.
【作者单位】： 中国农业大学水利与土木工程学院;
【基金】：国家重点研发计划项目(2016YFD0700204)
【分类号】：S831;TN912.3
本文编号：2167350
[Abstract]:Sound analysis technology has become an important tool for the study of animal behavior and animal welfare. The sound of laying hen can be used to evaluate its health or welfare status. However, there are many noise sources such as fan noise in the large chicken house, which has a great influence on the accurate identification of the sound of laying hens. The sound signal is used to reduce the noise of the fan. Using the digital sound acquisition platform to collect the sound of different types of laying hens and the noise and audio of the fan, Lab VIEW software is used to carry out the sound signal processing and analyze the time frequency characteristics of the sound and fan noise. At the same time, the different denoising methods (IIR filter denoising, wavelet threshold value) are compared and analyzed. Noise and improved spectral subtraction denoising effect in removing fan noise. The results show that, under the noise ratio of -8~20 D B, the improved spectral subtraction mean square root error is minimal (0.03~0.38), the algorithm running time is shortest (6~7 MS), and the effect of noise removal is better in practical application, and it can be the sound of the egg noise under the noise environment of the large-scale egg chicken house. Signal processing and analysis provide reference.
【作者单位】： 中国农业大学水利与土木工程学院;
【基金】：国家重点研发计划项目(2016YFD0700204)
【分类号】：S831;TN912.3
【相似文献】
相关期刊论文 前2条
1 顾锦生;钟翠珍;;饲料加工厂通风机噪声的治理[J];饲料机械;1982年02期
2 张彩霞;武佩;宣传忠;刘海洋;刘艳秋;;母羊声音信号处理与识别系统的设计[J];内蒙古农业大学学报(自然科学版);2013年05期
相关硕士学位论文 前3条
1 祁飞;基于Matlab的羔羊寻母声音信号的识别分析[D];内蒙古农业大学;2016年
2 张振华;基于音频分析技术的猪异常检测[D];太原理工大学;2017年
3 张彩霞;基于LabVIEW的母羊声音信号处理与识别系统研究[D];内蒙古农业大学;2014年
，本文编号：2167350
本文链接：https://www.wllwen.com/yixuelunwen/dongwuyixue/2167350.html
