| 摘要: |
| 可逆信息隐藏技术一方面能够对图像的原始性和完整性进行认证和保护，同时还能够确保无失真地恢复原始图像，近年来在公安、司法等领域受到越来越多的关注。基于可逆信息隐藏的认证方案需要同时满足可逆和认证两个方面的要求，在实际中具有较大挑战性，目前成功案例较少。在文献[1]中，Hong等人提出了一种新的基于可逆信息隐藏技术的认证方法，该方法借助IPVO （Improved pixel-value-ordering）和LSB （Least significant bit）替换等技术，可以有效地对图像进行认证。本文我们对Hong等人的方法进行了深入研究，指出在该方法中仅有部分像素参与认证码的生成且攻击方能够很容易地获知这部分像素，因此在安全性上还存在不足。针对该方法存在的安全漏洞，我们提出了一种针对性的攻击方案，即攻击方可选择对图像中未参与认证码生成和嵌入的像素进行修改。该攻击方案可以在不影响所嵌入认证码提取的同时，实现有意义篡改。为了提高认证算法的安全性，本文还针对Hong等人算法的缺陷提出了相应的改进方案，即将更多像素引入认证码的生成过程中并在嵌入前对图像块进行置乱。理论分析和实验结果验证了本文提出的攻击和改进方案的有效性。 |
| 关键词: 可逆信息隐藏 认证 IPVO LSB替换 |
| DOI：10.19363/J.cnki.cn10-1380/tn.2022.01.04 |
| 投稿时间：2021-04-02修订日期：2021-08-04 |
| 基金项目:本课题受国家自然科学基金（No.62072481，No.61772572）资助。 |
|
| Attack and Improvement of an Authentication Scheme Based on Reversible Data Hiding |
| WANG Hong,HUANG Fangjun |
| Guangdong Province Key Laboratory of Information Security Technology, School of Computer Science and Engineering, Sun Yat-Sen University, Guangzhou 510006, China;Guangdong Province Key Laboratory of Information Security Technology, School of Computer Science and Engineering, Sun Yat-Sen University, Guangzhou 510006, China;State Key Laboratory of Information Security, Institute of Information Engineering, Chinese Academy of Sciences, Beijing 100093, China |
| Abstract: |
| Reversible data hiding (RDH) technology can not only authenticate and protect the originality and integrity of the image, but also ensure that the original image can be restored without distortion. In recent years, it has attracted more and more attention in the fields of public security, justice and so on. Because the image authentication based on reversible data hiding needs to meet the requirements of reversibility and authentication function at the same time, it is quite challenging in practice, and there are few successful cases at present. In reference [1], Hong et al. proposed a new authentication scheme based on reversible data hiding. With the help of improved pixel-value-ordering (IPVO) method and the least significant bit (LSB) substitution technique, the image can be effectively authenticated without losing the reversibility. In this paper, the authentication method proposed by Hong et al. is deeply studied. We point out that in this method only part of pixels participate in the generation of authentication code, and the attackers can easily find them, so there are some deficiencies in security. Aiming at the security vulnerabilities of this method, a targeted attack scheme is proposed, that is, the attackers can choose to modify those pixels which are not involved in the generation of authentication code or the embedding process. Our attack scheme can achieve meaningful tampering without affecting the extraction of the embedded authentication code. Further, in order to improve the security of the reversible authentication algorithm proposed by Hong et al., in this paper we also propose a corresponding improvement scheme according to the defects of Hong's algorithm, that is, more pixels are introduced into the process of generating the authentication code, and besides, the Arnold scrambling is used before embedding the authentication code. Theoretical analysis and experimental results verify the effectiveness of the attack and improvement scheme proposed in this paper. |
| Key words: reversible data hiding authentication IPVO LSB substitution |
