基于频率转移的数字助听器单通道响度补偿算法  

A single-channel loudness compensation algorithm based on frequency transfer for digital hearing aids

作　　者：李战明[1] 张璇[1] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,甘肃兰州730050

出　　处：《电子设计工程》2017年第5期83-87,共5页Electronic Design Engineering

摘　　要：为了实现数字助听器响度补偿的目的,本文提出了一种基于频率转移的单通道响度补偿算法。该算法在频域分析的基础上,对语音信号高频部分按比例压缩并搬移到目标频段,再利用多分辨率小波对信号进行分解与重构,提取频谱包络,得到特征频率点初期信息;以特征点为端点进行分段,在各频段内进行响度补偿和增益控制。该算法使得补偿后语音的响度完全映射到患者的听觉范围内。仿真实验结果表明,该方法有效补偿了患者缺失的语音高频能量,显著提高了患者的听力水平及言语辨识率,同时保护了语音特征,达到了理想的效果。In order to achieve the purpose of loudness compensation,a single-channel loudness compensation algorithm is proposed based on frequency transfer technology for digital hearing aids .On the basis of frequency domain analysis for the algorithm, high frequency part of speech signal is compressed according to the proportion and transferred to the target band.And then the signal is decomposed and reconstructed by using the multi-resolution analysis of wavelet .Extract the speech spectral envelope, get the early information of frequency characteristics.Segmentation is performed by the feature points,then execute loudness compensation and gain control in each band.After loudness compensation,the identity of the signal is enhanced obviously and restricted within earshot of the hearing-impaired listeners. Simulation results indicate that the loss energy of the high frequency sound is well compensated, the level of hearing and speech recognition rate is improved, at the same time the speech feature is protected,which realized an ideal effect.

关 键 词：响度补偿 频率转移 多分辨率小波 特征频率点 

分 类 号：TN912[电子电信—通信与信息系统]