• •    

一种结合代数多重网格的钻石编码隐写算法

杨明1,黄颖1,2   

  1. 1. 重庆邮电大学
    2. 重庆市软件质量保证与测评工程技术研究中心
  • 收稿日期:2016-11-08 修回日期:2016-12-28 发布日期:2016-12-28
  • 通讯作者: 杨明

A novel image data hiding scheme based on algebraic multigrid and diamond encoding

  • Received:2016-11-08 Revised:2016-12-28 Online:2016-12-28
  • Contact: Ming YANG

摘要: 摘 要:通过代数多重网格方法(Algebraic Multigrid,AMG)将图像的像素点分成粗细网格两个部分,再结合钻石编码(Diamond Encoding,DE),把机密信息分别嵌入到粗细网格两个像素序列当中。粗网格部分像素的改变对整幅图像的质量影响较小,而细网格部分像素的改变对整幅图像的质量影响较大。又因为钻石编码中k值跟信息隐藏量密切相关,随着k值的增加像素改变量变大,所以用DE嵌入的过程中,粗网格部分选择较大的k值,细网格相反,在相同嵌入率的情况下较好地保证图像的感知度和安全性。实验使用峰值信噪比和一阶Markov安全指标进行衡量,通过与LSB置换、随机LSB匹配、DE算法与自适应边缘检测算法的比较,结合代数多重网格的钻石编码所提的三种方案都具有较高的安全性。

关键词: 数字隐写, 钻石编码, 代数多重网格, 图像结构相似性, Markov安全测度

Abstract: Abstract: The coarse grid part and fine grid part of an image are extracted with algebraic multigrid (AMG) method,and secret information is embedded into the two parts by diamond encoding (DE) method. The change of coarse part has little influence on the cover image quality, while pixels changing in fine part has a great effect on the cover image quality. The k values of diamond encoding are associated with the amount of information hiding closely, the pixel change a lot with the value of k increasing. Therefore in embedding process, a larger k values are chosen in the coarse part and a smaller k values are in the fine part. The experiments prove the proposed method have better perception and security compared with LSB replacement, random LSB matching, DE algorithm and adaptive edge detection algorithm with peak signal-to-noise ratio and first-order Markov safety index measure.

Key words: Keywords: digital steganography, diamond encoding, algebraic multigrid, structural similarity index, Markov security measure

中图分类号: