基于汉明码和同步信息的图像自嵌入算法(可编辑).doc

基于汉明码和同步信息的图像自嵌入算法(可编辑).doc

基于汉明码和同步信息的图像自嵌入算法(可编辑).doc

基于汉明码和同步信息的图像自嵌入算法  基于汉明码和同步信息的图像自嵌入算法# * 5 10 15 20 25 30 35 40 摘要:本文基于汉明码和同步信息提出一种图像自嵌入水印算法,可提高等分嵌入变长恢复 信息水印算法的恢复性能。该算法根据图像块的复杂程度生成变长恢复信息并将图像块的类 型信息进行(7,4)汉明码编码。按照图像块空间位置将变长恢复信息依次排列成一维序列, 选取同步图像块并对其恢复信息的起始位置编码生成同步信息。分别将类型编码信息与同步 信息组成的同步序列、恢复信息按照密钥随机置乱并等长嵌入到每个图像块。重构恢复信息 时,根据解码的类型信息,按照正序依次提取图像块的恢复信息。若图像块类型信息被篡改, 利用该图像块其后最近同步图像块的同步信息重构相应图像块的恢复信息。为提高图像的重 构率,本文再按照逆序利用同步信息重构相应图像块的恢复信息。实验结果表明,本文方法 提高了图像重构率,改善了算法的恢复性能。 关键词:变长恢复编码;等分嵌入;(7,4)汉明码;同步信息 中图分类号:TP309 Image Self-embedding Watermarking Scheme Based on Hanming Code and Synchronous Data ZHANG Ru, HE Hongjie, CHEN Fan School of Information Science and Technology,Southwest Jiaotong University, ChengDu 610031 Abstract: In this paper, a self-embedding watermarking algorithm is proposed based on the hamming code and synchronous data. The proposed scheme can improve the restoration performance of the watermarking scheme in which the alterable-capacity recovery data is embedded in the blocks with equal length. The proposed scheme generates alterable-capacity recovery data according to the complexity of blocks and encodes the type data of the blocks by 7,4 hamming coding method. The alterable-capacity recovery data of all the blocks are converted to a one-dimensional sequence. Then, the algorithm selects some synchronous blocks of which the starting position of recovery data is encoded to generate the synchronous data. The synchronous sequence, which consists of the encoded type data and synchronous data, and the recovery data are scrambled by the user key and embedded in each block with equal length respectively. In the process of recovery data reconstruction, the proposed algorithm extracts the recovery data in positive sequence with the decoded type data. When the type data of a block is destroyed, the synchronous data of the subsequent synchronous block is used to reconstruct the corresponding blocks. In order to improve the reconstruction rate, this paper also reconstructs the correspond

文档评论(0)

1亿VIP精品文档

相关文档