基于MATLAB的多通道人耳模型技术在皮肤听声中的应用  

Application of Twelve-channel Bandpass Filter in Skin-hearing Technology Based on MATLAB

作　　者：李建文[1] 李沙沙[1] 

机构地区：[1]陕西科技大学电气与信息工程学院,西安710021

出　　处：《计算机测量与控制》2012年第11期3083-3085,3088,共4页Computer Measurement &Control

基　　金：国家自然科学基金项目(60672001)

摘　　要：针对第一代皮肤听声器不能够准确进行语音辨析的问题,对人耳模型进行了多通道探索。根据中心频率和带宽设计了FIR滤波器,将300Hz～4kHz的音频范围内的音频信号划分为12个通道,音频信号经过12通道人耳模型的带通滤波,形成了电极片上的不同刺激;在MATLAB环境下实现了语音信号的12通道带通滤波及画出了相关语音的频谱图;仿真结果证明多通道人耳模型技术运用于皮肤听声系统中能够提高系统的稳定性,增强了辨析语音的能力。The twelve channel bandpass filter technology was proposed for the problem that the first generation of skin--hearing Aid can- not be accurate voice discrimination. FIR filter was designed according to the center frequency and the bandwidth, and the audio signal be- tween 400Hz and 5 kHz was divided into twelve channels, so the audio signals through the adoption of the twelve channels or cut off and the different stimulation was formed on the electrode. The voice signal twelve--channel band--pass filter was achieved and the relevant speech spectrum was drawled In the MATLAB environment. The results confirmed that the twelve bandpass filter can improve the system stability and the ability of Discrimination of voice was enhanced.

关 键 词：皮肤听声技术 多通道人耳模型 FIR滤波 MATLAB 语音辨析 

分 类 号：TP302.1[自动化与计算机技术—计算机系统结构]