Radon变换域的鲁棒图像Hash算法研究
Radon变换域的鲁棒图像Hash算法研究
张秀丽;雷延强;梁凡
【期刊名称】《计算机工程与应用》
hue trunc函数【年(卷),期】2011(047)016
【摘要】To solve image content authentication problem,a robust image Hash based on Radon Transform(RT) is presented.RT is performed on the input image, obtaining the invariant moment for translation and scaling, then Discrete Fourier Transform(DFT) is used to resist image rotation operation and reduce dimension of final Hash for image authentication.To enforce security, feature extraction is key-dependent.Experimental results show that the proposed algorithm can tolerate almost all content-preserving operations, such as scaling, rotation,JPEG compression, filtering and adding noises, while being sensitive to distinct distortion and essentially dissimilar image.%针对数字图像内容认证的问题,提出了一种
基于Radon变换域的鲁棒图像Hash算法.该算法在Radon变换域提取图像平移、缩放不变的矩特征,将矩特征的离散傅里叶变换系数作为图像的Hash值.离散傅里叶变换可实现抗图像旋转操作,并可降低图像Hash的维数.为了增加安全性,整个特征提取过程都是基于密钥进行的.理论分析证明Radon变换域的不变矩特征对一般的图像处理操作,特别是平移、缩放、旋转等几何操作有极好的不变性.实验结果表明提出的算法几乎可以容忍所有的图像内容保存操作,比如缩放、旋转、JPEG压缩、滤波和加噪等,同时对视觉上有明显失真或根本不相似的图像有高度的敏感性.

版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。