| Digital hearing aids have great results in helping solve hearing loss of the hearing-impaired patients. This article has further research on speech enhancement technology and loudness compensation algorithm in digital hearing aids. Traditional algorithms are simulated, and new algorithms are proposed. They are aimed at improving the performance of digital hearing aids. The main achievements of the work are listed as follows:Firstly, the paper studies the mathematical model of the generation of speech signal and the key technologies commonly used in digital hearing aids, including source localization, acoustic echo cancellation, and wide dynamic range compression. This article focuses on the speech enhancement technology and loudness compensation algorithm.Secondly, the paper studies the speech enhancement based on spectral subtraction, improves the traditional spectral subtraction by setting the correction factor and raise the effectiveness of de-noising. Emphatically analyses the speech enhancement based on wavelet threshold de-noising,puts forward a new threshold function. The performance of de-noising is improved, and the proposed threshold function is good in the function of switch selection. By adjusting the parameters,the improved threshold function can transform from one function to another, and this will lead to different performance.Thirdly, the paper has a further research on the loudness compensation algorithm which is commonly used in digital hearing aids. On the basis of human auditory characteristics, describes the principles and commonly used algorithms of loudness compensation. This paper puts forward a new method of nonlinear loudness compensation. It does not need to divide the frequency bands and avoid the waveform distortion which may caused by frequency division. The new algorithm can adjust the variation of compensation factor according to the hearing-impaired listeners’ real-time awareness to sound intensity and improves the effect of loudness compensation. |
PDF Full Text Request