کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
537162 | 870709 | 2007 | 9 صفحه PDF | دانلود رایگان |
![عکس صفحه اول مقاله: A power-based unequal error protection system for digital cinema broadcasting over wireless channels A power-based unequal error protection system for digital cinema broadcasting over wireless channels](/preview/png/537162.png)
This paper deals with the efficient and robust wireless broadcasting of JPEG2000 digital cinema (DC) streams from studios to theatres. Several unequal error protection (UEP) techniques have been proposed in literature for the transmission of JPEG2000 images. Some are based on variable forward error correction (FEC) coding applied to different parts of the stream according to their importance. Alternatively, UEP can be achieved by means of unequal power allocation (UPA) schemes based on differentiated transmission power over the stream. On the other hand, in DC applications UPA achieves weak performance if considered as the only protection strategy, unless high-power budget is assigned to transmission. This work proposes a novel hybrid FEC–UPA system adopting the resilience tools of the JPEG2000 wireless (JPWL) standard. The JPWL stream is partitioned into a certain number of packet groups to which “light” FEC coding is applied. Groups are then transmitted through separate wavelet packet division multiplexing (WPDM) sub-channels at different power. Both stream partitioning and UPA are driven by the sensibilities of the JPWL packets to the channel errors. The novelty of the proposed technique relies on the use of JPWL error resilience tools for the codestream partitioning, on optimized UPA among JPWL packets based on genetic algorithms (GA) and supported by “light” FEC channel coding. The proposed system is compared to the state of the art UEP techniques on JPEG2000 transmission. The performance is evaluated in case of transmission over wireless channels with both sparse and packet error statistics. Experiments show that the proposed approach allows achieving an average peak signal-to-noise ratio (PSNR) on the reconstructed frames compliant to the standard quality required by DC applications (40 dB) for bit error rate (BER) up to 10−4.
Journal: Signal Processing: Image Communication - Volume 22, Issue 3, March 2007, Pages 331–339