Share Email Print
cover

Proceedings Paper

Progressive diversity allocation for progressive transmission of spiht packets in erasure channels
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The SPIHT (set partitioning in hierarchical trees) algorithm, which works with the aid of the discrete wavelet transform (DWT), is one of the most renowned techniques for image compression. One problem with SPIHT image coding, especially when image transmission is involved, is that even just as few as one error in the embedded code sequence can render the decoded image completely unrecognizable. In this paper, we propose a scheme in which an SPIHT code sequence is packetized into two types: CP (critical packet) and RP (refinement packet). Then, we address the transmission of those SPIHT packets over erasure channels. The basic idea is to provide unequal error protection (UEP) to those packets by diversity (i.e., repeated transmissions) so that, at the receiver, the decoded image quality (measured in the expected SNR, signal-to-noise-ratio) is as good as possible. A diversity allocation (DA) algorithm, referred to as progressive diversity allocation (PDA), is proposed. It works in a fashion and is naturally compatible with the progressive transmission of SPIHT-encoded images. Experiments show that the PDA scheme produces good results, nearly as good as achieved by the method of full search.

Paper Details

Date Published: 14 March 2013
PDF: 7 pages
Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87687G (14 March 2013); doi: 10.1117/12.2010075
Show Author Affiliations
Kuen-Tsair Lay, National Taiwan Univ. of Science and Technology (Taiwan)
Lee-Jyi Wang, National Taiwan Univ. of Science and Technology (Taiwan)
Tungnan Univ. (Taiwan)


Published in SPIE Proceedings Vol. 8768:
International Conference on Graphic and Image Processing (ICGIP 2012)
Zeng Zhu, Editor(s)

© SPIE. Terms of Use
Back to Top
PREMIUM CONTENT
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?
close_icon_gray