摘要
图像多小波系数在邻域内具有相关性,调整相关性大的系数会对图像视觉效果产生影响,为了减小这种影响,提出一种基于灰色关联分析的多小波域图像隐写方法。该方法运用灰色关联分析对邻域内多小波树组的中心树序列和外围树序列的几何曲线形状进行比较,描述它们之间的相关性,从而在矩阵中找出与相邻点相关性很小的系数树作为秘密信息的嵌入位置。结合多小波变换系数能量更分散的特点,使得含密图像的质量更好。实验结果表明,该隐写算法具有更好的隐蔽性和鲁棒性。
Muhiwavelet transform coefficients of image have relativity in adjacent region. Adjusting coefficient which has large relativity can bring influence on vision effect of image. In order to reduce the influence, the paper proposed a steganography of image in muhivavelet domain based on Grey Relational Analysis （ GRA）. The steganography used GRA to compare geometric curve shape of central tree sequence and peripheral tree sequence of multiwavelet tree group in adjacent region, depicted their relativity, consequently found out coefficients tree that has very little relativity with other adjacent point in coefficient matrix as embedded location of secret information. And it made the quality of stegano image better by utilizing more dispersive characteristic of muhiwavelet transform coefficient. better imperceptibility and robustness.
出处
《计算机应用》
CSCD
北大核心
2009年第12期3182-3184,共3页
journal of Computer Applications
关键词
数字隐写
灰色关联分析
多小波
视觉特性
steganography
Grey Relational Analysis （GRA）
Experimental result indicates the proposed algorithm has muhiwavelet
Human Vision System （HVS）
