Research And DSP Implementation On Frequency Response Compensation Of Digital Hearing Aid In Noise Environment

Hearing impairment seriously affects people's normal life,long-term hearing impairment not only affects people's daily communication,but also induces mental illness.Wearing digital hearing aids can effectively improve the hearing level of hearing-impaired patients.Speech enhancement and frequency response compensation are two important digital signal processing techniques in digital hearing aids that directly affect the performance of digital hearing aids.In addition,the development of digital signal processing technology provides a good platform for the implementation of algorithms.Therefore,this paper mainly improves the frequency response compensation algorithm in the noise environment and implements the improved algorithm in real time on the DSP hardware platform.The research work in this paper mainly includes the following two parts:First,the existence of noise seriously reduces the frequency response compensation effect of digital hearing aids.In order to improve the speech recognition rate after frequency response compensation,this paper uses binary masked speech enhancement algorithm for denoising before frequency response compensation.Firstly,the algorithm uses the minimum control recursive averaging algorithm to estimate the noise of noisy speech,then perform Wiener filtering on the noisy speech to initially remove the noise,finally,the sub-band decomposition of the preliminary enhanced speech is performed using a Gammatone filter and binary masking is performed in each sub-band,noise is effectively removed,reducing speech distortion.Secondly,the improvement of speech intelligibility of digital hearing aids in noisy environments does not meet the needs of hearing-impaired patients,therefore,this paper proposes a multi-channel adaptive frequency response compensation algorithm based on signal-to-noise ratio in noisy environment.The algorithm first calculates the sub-band signal-to-noise ratio and frequency response compensation gain of the speech after enhancement by binary masking,Then,using the advantages of linearization and wide dynamic compression compensation under different signal-to-noise ratios,the target speech is compensated adaptively according to the signal-to-noise ratio of the sub-band and the hearing curve of the hearing-impaired patient.Experiments show that the proposed algorithm takes noise into account in thefrequency response compensation,it protects the formant structure and spectral envelope of speech,the sound pressure level of speech is more compensated in the hearing comfort threshold of hearing-impaired patients.Thirdly,the improved algorithm is transplanted to the development board based on TMS320C6416 DSP chip,and the algorithm is realized by the connection between the various components of the hardware system and the initial configuration of the main modules in the system.The test shows that the proposed algorithm can be implemented in the DSP system,and the speech processed by system meets the hearing needs of the hearing-impaired patients in the noisy environment.