Speech compression using discrete wavelet transform with optimum differential pulse code modulation (DPCM)‎

Abstract EN

This paper the performance of speech compression system using discrete wavelet transform (DWT) is investigated.

Two methods are used for this purpose; in the first only discrete wavelet transform is used.

In the second method linear predictive coding (LPC) in wavelet transform is used.

Both methods are used differential pulse code modulation (DPCM) for quantization with optimum selection of quantization parameters using Lloyd algorithm.

Huffman code is used for both methods to increase the compression ratio without effect on the quality performance of speech signal.

The results show that DPCM with DWT gives the best performance than when it is used with LPC and DWT.

For speech compression using DWT, for n = 3 and 4, DblO gives high quality measure and less CF.

For n = l and 2, Db2 gives better results for compression factor and quality measures.

For speech compression using LPC and DWT, Db6 gives good performance.