Algorithm Of Speech Enhancement Based On Wavelet Transform And Implementation Based On DSP

Speech enhancement is an effective method to solve the noise pollution, it has wide application areas in communication, network, speech recognition, speech coding, and so on. In recent years, with the development of VLSI technology and high-speed DSP chips has made speech enhancement of real-time realization become possible speech enhancement method progressively toward practical, people has been paid more attention to speech enhancement technology.This paper firstly introduces the purposes, application, development history, and status of speech enhancement, This paper analyses characteristics of speech and noise, then discussed the basic principle of the several common speech enhancement, such as the Wiener filter, Kalman filter, spectral subtraction, adaptive filtering, short-term spectral amplitude estimation minimum mean square error, and analyze their advantages and disadvantages.Discusses the wavelet threshold denoising method, and the basic principle of the threshold value rules, To simulate the threshold selection, the threshold function, decomposition layers and selection of wavelet radix, and select the optimum parameters; In the study of traditional wavelet threshold denoising algorithm, and the advantages and disadvantages of traditional threshold algorithm were analyzed. Then in view of the disadvantages, function was improved, and the simulation results show that the improved threshold function de-noising result is superior to the traditional threshold function; Proposed a method based on wavelet threshold, combined with improved spectral subtraction algorithm, simulation results show that: the method can effectively enhance by additive white noise and noise on speaker's voice, effectively reduce the music noise.Design a the speech enhancement module based TMS320VC5416 fixed-point DSP, discusses the theory of the hardware circuit design process modules. This paper based on the improved algorithm design the system and algorithm programming , and realizes the algorithm on CCS and runs on the system successfully on the ICETEK-VC5416-A.