| 摘要 |
<p>A method and system is provided for encoding and decoding of speech signals at a low bit rate. The continuous input speech (15) is divided into voiced and unvoiced time segments of a predetermined length. The encoder of the system (100) uses a linear predictive coding model for the unvoiced speech segments and harmonic frequencies decomposition for the voiced speech segment. Only the harmonic frequencies are determined using the discrete fourier transform of the voiced speech segments. The decoder (400) synthesizes voice speech segments using the magnitudes of the transmitted harmonics and estimates the phase of each harmonic from the signal in the preceeding speech segments. Unvoiced speech segments are synthesized using linear prediction coding coefficients obtained from codebook entries for the poles of the LPC coefficient polynomial. Boundary conditions between voiced and unvoiced segments are established to insure amplitude and phase continuity for improved output speech quality.</p> |
