Compression of Psuedo-Periodic Signals Using 2D Wavelet Transforms

This paper attempts to utilize the pitch synchronous property of pseudo- periodic 1-dimensional signals like voiced speech, music etc, to improve the efficiency of compression retaining the distinctive characteristics of the original signal. The signal is represented in 2-dimensional form and decomposed using 2-D wavelet. The decomposed signal is compressed using various threshold parameters. Results show that higher signal to noise ratio, higher compression ratio and lower percentage distortion are obtained with the new method of 2-D compression as compared to 1-D compression. We have used a new method of pitch peak detection of voiced signals using k-means clustering algorithm.