项目名称: 基于随机但可估计嵌入的安全数字水印算法
项目编号: No.61300208
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 曹坚
作者单位: 哈尔滨工业大学
项目金额: 25万元
中文摘要: 数字水印技术不但需要能抵抗常规的信号处理操作(即要有强的鲁棒性)而且还需要能抵抗密钥估计(即要有强的安全性)。否则,它将不可用于像版权保护和盗版跟踪之类的水印应用。然而,现有的大多数数字水印算法都不能抵抗密钥估计攻击。尽管近几年一些安全的数字水印算法已经被提出,然而这些水印算法都有鲁棒性不强的缺陷。本项目主要研究如何随机化水印的嵌入方向或者随机化扩频的嵌入子空间的正交基,以使得攻击者不能(但解码者能)估计它们,来设计既安全又鲁棒的数字水印算法。主要研究内容包括:使用传输理论和递归思想来减少旋转不变水印算法的嵌入失真;在嵌入子空间使用随机但可估计的正交码本来设计旋转不变水印算法;扩展变换抖动调制改进:可抗未经授权去除攻击且缩放攻击;基于随机但可估计嵌入方向的安全水印算法。本项目突破了仅基于固定检测边界和扩频技术来构建安全水印算法的现有框架,为设计既鲁棒又安全的水印算法提供了新的思路和方法。
中文关键词: 版权保护;数字水印;鲁棒性;安全性;扩频
英文摘要: Digital watermarking technique should resist both common signal processing operations (namely, it must have strong robustness) and the estimation of secret key (namely, it must have strong security). Otherwise, it is not applicable to the watermarking applications such as traitor tracing and copyright protection. However, most existing digital watermarking algorithms cannot resist the estimation of secret key. In recent years, some secure digital watermarking algorithms have been proposed. However, these secure digital watermarking algorithms aren't very robustness. The project mainly researches how to design both secure and robust digital watermarking algorithms by means of random embedding direction or random orthogonal base of embedding subspace in a sense that the attacker cann't (but the decoder can) estimate them. Main research contents include: reduce the embedding distortion of circular watermarking by means of recursive idea and transportation theory; use random but estimable and orthogonal codebook for designing circular watermarking in the embedding subspace; improved spread transform dither modulation for assisting both unauthorized removal attack and scaling attack; secure watermarking algorithm based random but estimable embedding direction. This project breaks through the existing framework wh
英文关键词: Copy protection;digital watermarking;robustness;security;spread spectrum