解碼端影像誤差估測用於分散式視訊編碼法校正優先權的設計

Abstract

本篇論文提出了一種使用於分散式視訊編碼的技術，對於W-Z frame提供了有優先順序的channel coding。在我們所提出的架構中，基於預估的side information誤差程度，W-Z frame的macro-block會被分類為幾個不同的群組。分類的資訊經由一個上傳channel傳回編碼端，所以誤差程度相似的macro-block會被聚集在一起以進行channel coding。使用這種方法，解碼端可以為side information品質比較不好的macro-block要求多一點的parity bits以更正這些部分的錯誤。而對於side information誤差較小的macro-block則要求比較少的parity bits。比起目前最好的DISCOVER DVC codec，對於QCIF sequences，在較低的bitrate範圍裡(低於200 kbps)我們所提出的DVC架構可以使R-D performance增加0.3至0.5 dB。因為低複雜度的編碼器對於像是sensor network之類的低bitrate監視應用系統是很重要的，所以我們所提出的架構可望應用到實際系統上。

In this thesis, a distributed video coding technique with prioritized channel coding of W-Z frames is presented. In the proposed framework, W-Z frame macro-blocks are classified into different groups based on the estimated errors of the side information. The information is transmitted via uplink channel back to the encoder so that macroblocks with similar error statistics can be grouped tighter in same coding blocks for channel coding. With this approach, decoder can request more parity bits to correct macroblocks whose side information quality is worse and request less or no parity bits for macroblocks with small side information errors. The proposed DVC scheme can increase R-D performance about 0.3 to 0.5 dB over the state-of-the-art DISCOVER DVC codec for low bitrate (less than 200 kbps) applications for QCIF sequences. Since low-complexity encoder is important for low bitrate surveillance applications such as those for sensor networks, the proposed scheme is very promising for practical applications.