電子透かしアルゴリズムの安全性評価手法とバイオメトリクスへの応用に関する研究

在这项研究中，我们提出了一种新的语音水印方法，通过在语音产生过程中操纵清晰度。该方法是通过建立与语音产生过程等价的准声道模型来实现的。水印声道模型用LSP(Line Spectrum Pair)参数构成的码块表示。语音水印的过程如下：1)从语音中提取LSP；2)将提取的LSP中的一部分替换为码本向量。3)使用替换的LSP合成语音。在此过程中，将水印间接地嵌入到语音质量和准确度测试中，并将该方案应用于ITU-T G.729(CS-ACELP)中，期望数字水印对各种篡改具有鲁棒性。然而，许多水印不一定对所有类型篡改具有稳健性。为了避免这一问题，我们提出了一种混合型数字水印。本文将提出的数字水印方法应用于指纹识别系统中的指纹识别系统中，指纹数据的泄漏是一个严重的问题，用户很难将泄漏的原因归咎于管理员，因为用户无法证明图像是从哪里泄漏出来的。

In this study, we propose a new watermarking method for speech by manipulating the articulation in the process of speech production. The proposed method is performed by modeling a quasi vocal tract model equivalent to the speech production process. The watermarked vocal tract model is expressed by codebocks made by LSP(Line Spectrum Pair) parameters. The procedure of watermark for speech is as follows ; 1) LSPs are extracted from the speech 2) Some of the extracted LSPs are replaced by the codebook vectors. 3) Speech is synthesized using replaced LSPs. In the process above, watermarks are embedded indirectly into the speech Evaluation tests on speech quality and accuracy of the proposed method a scheme is applied to the ITU-T G. 729 (CS-ACELP).Digital watermark is expected to have robustness against various alterations. However, many watermark do not necessarily have robustness against all kinds of alterations. To avid this problem, we propose a hybrid type of digital watermark. The hybrid type of digital watermark is rather robust than the conventional ones by using some digital watermarks that have robustness against different kinds of alterations.Furthermore, we apply our proposed watermark method to a fingerprint identification system In a fingerprint identification system a leak of a fingerprint data is a serious problem Is is difficult for a user to blame administrators for the leak, because the user cannot prove where the images leaked out from We propose a fingerprint identification system that enables a user to prove it by using digital watermarks.