| Summary: | 碩士 === 國立臺灣科技大學 === 電子工程系 === 99 === Due to the rapid development of multimedia technologies, the effective compression and transmission methods against the limited storage space is needed. The characteristics of wavelet transform make it suitable for image compression. Coupled with the discrete wavelet transform, set partitioning in hierarchical trees (SPIHT) is a highly efficient image compression technique that allows for progressive transmission. One issue, however, is that its decoding can be extremely sensitive to errors in the embedded bit stream generated by the SPIHT encoder. The transmission of images over noisy channels is still a challenging problem.
In this thesis, we address the issue of transmitting SPIHT-encoded images via
noisy channels, wherein errors are inevitably encountered. The communication scenario assumed in this thesis is that no acknowledge path is available from the receiver to the transmitter, and thus error protection has to be done by forward error correction.
In our scheme, the original SPIHT code sequence is packetized. The resulted packets are then allocated with different diversity orders for unequal error protection (UEP) so that the decoded image quality is optimal (in the sense of the expected SNR) at the receiver. Experimental results show that the proposed scheme significantly improves the quality of the decoded images as compared to the case of no diversity allocation. Moreover, comparisons with existing UEP techniques for transmitting SPIHT-encoded images are made. The results show that our method outperforms those existing UEP techniques.
|
|---|
