摘要
基于同态加密和双层隐写编码,该文提出一种安全的密文域图像隐写术,其可以达到传统明文隐写术的容量,并且在密文域和明文域均能有效抵抗隐写检测分析。首先结合自适应隐写术和湿纸编码技术,提出一种明文域双层隐写算法;其次,修正一种全同态加密算法,对载密图像进行加密;最后,在密文域上提取嵌入的信息。理论分析和实验结果表明:在加密/隐写密钥同时泄露、加密密钥泄露和密钥未泄露条件下,算法均具有较高的安全性。
A new secure image steganographic algorithm for encrypted domain steganography is proposed based on fully homomorphic encryption and double-layer embedding. The proposed method can reach the capacity of classical steganography while resist to statistical steganalysis both in plaintext domain and encrypted domain. At first, a novel double-layered embedding algorithm based on adaptive steganography and Wet Paper Codes（WPCs） is constructed for plaintext embedding. Then an fully homomorphic encryption algorithm is modified to encrypt the stego image. Finally, the embedding message is extracted in encrypted domain. The performance analysis and experimental results show that the steganographic security of the proposed method preserves well performance under several attacking conditions, including both steganographic/encrypted keys leakage, encrypted keys leakage and no keys leakage.
出处
《电子与信息学报》
EI
CSCD
北大核心
2012年第7期1721-1726,共6页
Journal of Electronics & Information Technology
基金
国家自然科学基金(60803155
61170234
60970141
60902102)
郑州市科技创新团队项目(10CXTD150)资助课题
关键词
隐写术
自适应编码
湿纸编码
同态加密
Steganography
Adaptive coding
Wet Paper Coding （WPC）
Homomorphic encryption
