Sistema de redução de ruído de baixo custo para sistemas de auxílio à audição: algoritmo RACE.

Abstract

Hearing impairment affects about 1.5 billion people worldwide,10.7 million of which only in Brazil. As a solution to maximizehearing capacity, the use of hearing aids is commonly indicated.However, the adherence of these devices is low, largely due to thelow sound quality in noisy environments and their high cost. Tominimize these obstacles, several studies and research are carriedout both on the hardware part and on noise reduction techniques.The technique addressed in this work is the Real Time AdaptativeCorrelation Enhancer algorithm, an adaptive filter that uses theautocorrelation coefficients of the noisy audio signal to calculatethe coefficients of Wiener’s FIR (Finite Impulse Response) filter. Aspreliminary results, the simulation of the algorithm in MATLAB®ispresented to discuss the performance of the algorithm. As futurestudies, it is planned to implement the algorithm on the TexasInstruments®MSP432 microcontroller low-cost development board,to analyze the algorithm’s performance with samples in real time.