Speech Processing Algorithm Base On Auditive Perception For Cochlear Implant

Cochlear implant is the only effective equipment has the ability to renew auditive system of deaf people. Simulating the physiological function of normal people's auditive system on concha, eardrum, inner ear is the basic function of cochleat implant. Inner ear is the key in the auditive system. Speech signal processor is the key of the cochlear implant, influencing the capability of cochlear implant directly. Cochlear implant become the focus and hotspot to research.In this thesis speech signal processing is mainly discussed, proposing the improvement speech signal processor arithmetic. Proposing the auditive perception principle applying to the speech signal processing based on the wavelet transform, combining the auditive perception character of speech. The recognizing ability of cochlear implant is improved after simulating. The main work of this thesis is below:1. Proposing the auditive perception wavelet transform. Adopting the multi- distinguish analysis and dividing the time-frequency space non-equably, made the relationship of auditive perception region and length of eardrum linearity and logarithmic between auditive perception region and speech frequency. The speech signal after processing is more fit for the human auditive system.2. Research and discuss all kind of cochlear implant speech signal processing precept. Applying the auditive perception into cochlear implant speech signal processing precept. The new method on the basis of auditive perception was presented to easily adjust the parameters of filter bank and realize fast algorithm of speech processing for cochlear implants.3. Research the character of speech signal: format and pitch. Applying the auditive perception character into auditive perception cochlear implant speech signal processing. Reserving the character of speech signal and improving the recognizing ability.4. Carrying on simulating the cochlear implant speech signal processing. Presenting the results and discussion.In this thesis, proposing a bran-new algorithm on cochlear implant speech signal processing. Provide a platform for researching new algorithm on cochlear implant speech signal processing.