ANAUDIO DENOISING APPROACH USING HYBRID MODIFIED FAST WAVELET TRANSFORM METHOD

Abstract

Noise includes all unwelcome ambient sounds. If a sound is perceived as delightful music or annoying noise, it will be perceived as having the same decibel level. The fundamental drawback of noise in an audio signal is that it degrades the signal's quality during transmission over the communication system. Real-time audio signals are gathered for the present study using a microphone. The additive white Gaussian noise of AWGN is blended with a genuine audio source. The Median, Finite Impulse Response filter with Wavelet Transform approach is a hybrid filter for denoising audio signals that incorporates the Median Filter, FIR Filter, and Wavelet Transform techniques (MFWT). In this study Peak Signal to Noise Ratio and Mean Square Error (MSE) are used to assess MFWT performance(PSNR). In terms of SNR (82.04) and RMSE score, MFWT performs better than the median filter, FIR filter, and Wavelet transform (0.055).