ロスレス圧縮を用いて証拠音声データを対象とした改ざん検出可能な可逆ステガノグラフィ(マルチメディア通信/システム,ライフログ活用技術、IP放送/映像伝送,メディアセキュリティ,一般)

概要

論文の詳細を見る
A reversible and verifiable tampering detection steganographic method is described for probative acoustic data that uses lossless compression with the Huffman algorithm partially applied to the sampled data stream in the perceptual psychoacoustic time domain. Speech data is divided into fixed-size frames, and the feature value of each frame is calculated, embedded in the capacity location reserved by the compression algorithm, and then used to detect tampering and identify the tampering positions in each frame. Embedding and extraction algorithms are used to achieve reversibility and are implemented using speech data. Objective testing shows that the proposed method is effective and that the distortion meets ITU-T recommendations.
2011-09-13