資料隱藏技術除了將大量機密資料嵌入數位影像之外,也必須確保在接收端所重建掩護影像之可逆性。本篇論文提出一種實現於邊緣匹配向量量化影像壓縮領域之「可逆式而且高負載之資料隱藏演算法」。所提方法使用機密資料藏入至影像之邊緣匹配向量量化壓縮碼的策略,來實作資料隱藏技術。所提方法能將壓縮碼重建出原始向量量化壓縮影像達到可逆式的效果,進而強化向量量化壓縮技術之實用性。除此之外,霍夫曼編碼技術也進一步壓縮接收時的整體資料量,如此將有利於提高資料嵌入量。最後,藉由實驗數據分析發現,本論文提出之方法除了有效地維持原始量向量化壓縮掩護影像品質之外,再與其它演算法比較時有較低的平均位元率與較高的平均嵌入率。
The data hiding technology not only embed the high-payload secret data into digital image, but also could reconstruct the original cover image at the receiver. “A reversible and high-payload data hiding scheme implemented in the SMVQ compression domain of image” is proposed in this thesis. The idea of this scheme is to hide secret data into the compression codes of image by utilizing the sorted state codebook of SMVQ. The compression codes of image reversibly reconstructed original VQ-compressed cover image in the proposed scheme. Besides, the Huffman coding technique is applied to compact the volume of the overall data needed to be transmitted. The proposed scheme significantly enhances VQ-compressed technique and achieves high embedding capacity. Eventually, the experimental results show that the proposed scheme maintains good visual quality for the reconstructed original VQ-compressed cover image. Moreover, the proposed scheme achieves the best performance among approach in literature with the average low bit rate and high embedding rate.