International Journal of Computer Science Issues

More than a traditional journal...

When the speech signal corrupted by noise is processed using different transforms like the Discrete Fourier Transform, Discrete Cosine Transform and the Discrete Sine Transform , a comparative analysis proves that the Discrete Sine Transform (DST) is most suitable for de-noising and therefore reconstruction of the original speech. The experimental results show that the discrete sine transform coefficients at the low frequency regions are predominantly speech, and at the high frequency regions are predominantly noise. According to this, a new noise estimation method based discrete sine transform is proposed in this paper. The usage of DST coefficients reveals that the energy distribution throughout the segment of speech is uniform.