I am working on transmitting compressed images over wireless fading channels. When I used the SPIHT-based compression (using dyadic wavelet decomposition), I noticed that the sensitivity to bit error is too large and the quality of the decoded image at the receiver is low. However, when I applied SPIHT on wavelet packet, the senstivity to bit error decrease noticeably, and the qualitY of the decoded image is improved noticeably. Could you please help me interpreting this result.
The results when wavelet packets are employed are better than when wavelets are applied. This is quite obvious, because all the subbands are explored in the wavelet packet case. And the thing that you need to note is that the transforming time as well as coding and decoding time will be higher than that with wavelets. And the amount of time that you spent is a limitation. The second parameter is that the processing complexity with wavelet packet is about 4 times that of wavelets. This results in coding and decoding algorithms also.
Dear Asmaa,
You have to mention the degree of compression in the two cases. The transform coding is a lossy compression and in order to judge the two methods one has to see the resulting degree of compression in addition to the picture quality after decoding that is peak pixel signal to noise ratio with an ideal channel.
More lossy compression will lead to more signal to noise ratio.
Best wishes
- Badges
- Science topic