SPEECH COMPRESSION USING BEST WAVELET PACKET TRANSFORM AND SPIHT ALGORITHM

Abstract

Speech compression or Speech coding is the process of converting human speech into an efficiently encoded representation that can later be decoded to produce a close approximation of the original signal. The current work considers the application of discrete Wavelet packet transform, for the compression of human speech signals. Wavelet packet analysis is the breaking up of a signal into a set of scaled and translated versions of an original (or mother) wavelet. Taking the wavelet transform of a signal decomposes the original signal into wavelets coefficients at different scales and positions. The task of encoding these wavelet coefficients is carried out by Set Partitioning in Hierarchical Trees (SPIHT) algorithm which is widely used in still image coding. The Set Partitioning in Hierarchical Trees (SPIHT) algorithm is a coding algorithm that allows the transmission of coefficients in a magnitude sorted fashion where the most significant bits of the largest coefficients are sent first. It has been found that by using this approach, we can reconstruct the speech with appreciable perceptible quality. The Matlab code is written and simulated for male, female and recorded speech test signal samples. It . is observed that for male speech samples and recorded speech Biorthogonal family of wavelets give high SNR and for female samples Daubachies wavelets give high SNR for all the samples.