摘要
提出一种EZW编码图像中的安全隐写算法,该算法以嵌入前后EZW重构小波系数分布的相对熵作为安全指标,采用熵解码变换的方法,将数据比特映射为具有指定分布的0/1比特,并利用量化嵌入机制实现数据的嵌入.所提算法可实现载体和隐写信号一阶统计特性的保持,因此具有一阶的统计安全性.实现结果同时表明,该算法在容量、感知质量和安全性上具有较好的综合性能.
A secure steganographical algorithm is proposed.The algorithm takes the relative entropy of the EZW(embedded zerofree wavelet)-reconstructed wavelet coefficient distributions before and after data embedding as the security object,and uses the entropy decoding transformation method to map the data bits to the bit sequence with given distribution and perform data embedding based on quantization.The proposed method can hold the first-order statistics,so it has the first-order security.The experimental results also show that the proposed algorithm has better synthetic performance in capacity,perceptibility quality and security.
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2008年第6期975-980,共6页
Journal of Southeast University:Natural Science Edition
基金
国家自然科学基金资助项目(60374066).
关键词
隐写
EZW
相对熵
量化嵌入
steganography
embedded zerofree wavelet
relative entropy
quantization-based embedding