Subband- and Wavelet-Based Coding - Digital Signal Processing: Fundamentals and Applications - page 681

a. Use a 16-bit code for each wavelet coefficient and write a MATLAB program to perform
data compression with the following ratios: 2:1, 4:1, 8:1, 16:1, and 32:1.
b. Measure the SNR in dB for each case.
c. Plot the reconstructed waveform for each case.

13.10.2 MATLAB Projects
13.32. Data compression using subband coding:
Given 16-bit speech data (“speech.dat”) and using the four-band subband coding method,
write a MATLAB program to compress a speech signal with the following specifications:
a. 16 bits for each of the subband coefficients, code the LL, LH, HL, HH subbands, and
measure the SNR in dB.
b. 16 bits for each of the subband coefficients, code the LL band, discard the LH, HL, and HH
subbands.
c. 16 bits for each of the subband coefficients, code the LL and LH bands, discard the HL and
HH subbands.
d. 16 bits for each of subband coefficients, code the LL, LH, and HL bands, discard the HH
subband.
e. Measure SNR in dB for (a), (b), (c), and (d).
f. Determine the achieved compression ratios for (a), (b), (c), and (d).
g. Repeat (a) to (f) for seismic data (“seismic.dat”) in which each sample is encoded using 32
bits instead of 16 bits.

13.33. Wavelet-based data compression:
Given 16-bit speech data (“speech.dat”) and using the wavelet coding method with 16-tap
Daubechies wavelet filters, write a MATLAB program to compress a speech signal with
the following specifications:
a. 16 bits for each of the wavelet coefficients, compression 2:1.
b. 16 bits for each of the wavelet coefficients, compression 4:1.
c. 16 bits for each of the wavelet coefficients, compression 8:1.
d. 16 bits for each of the wavelet coefficients, compression 16:1.
e. 16 bits for each of the wavelet coefficients, compression 32:1.
f. Measure SNR in dB for (a), (b), (c), (d) and (e).
g. Repeat (a) to (f) for seismic data (“seismic.dat”) in which each sample is encoded using 32
bits instead of 16 bits.