在不良的網路環境下傳輸區塊編碼影像，容易造成影像區塊的遺失。為了不增加因重新傳輸資料造成網路的負荷且有效重建受損的影像，我們提出一個基於小波維度分析為基礎的即時影像重建技術。我們所提出的影像修補技術，先把影像依據周圍紋理的複雜度，分成平滑和非平滑兩類。若是破損區塊周圍為平滑的，則使用周圍完好像素的平均值修補破損像素。若是此破損區塊周圍為非平滑的，則在小波LL2區域中的破損區塊，利用候選窗與參考窗的均方差來找尋相似度最高的區塊作重建，而其餘頻帶的破損則根據水平垂直對角的紋理資訊來決定候選窗的範圍，並決定相似度最高的區塊，來修補其餘的頻帶。將修補完之所有係數做反小波合成，得到完成修補後之影像。我們使用不同程度破壞的影像作為測試影像，並且與其他的即時重建的演算法做比較，證實其有效性。 In the transmission of block-coded images , loss of a single bit often results in the loss of a whole block. Instead of using common retransmission query protocols , a reconstruction method based on multi-resolution analysis in wavelet domain is proposed. The first step in reconstruction is to classify missing image blocks into smooth and non-smooth ones. If the lost block is smooth , it is reconstructed using the available information surrounding it. If a block is classified as non-smooth one, the multi-resolution characteristics in the wavelet transform, from the lowest spatial-frequency layer to the higher one, to analyze the image progressively from global-area to local-area. The proposed algorithm makes use of the wavelet-domain information of not only neighboring regions , but also remote regions within the image. The performance of this method is tested for the test images.The experiment as results show that the proposed algorithm can provide high quality reconstruction of images.
