1. 吉林大学计算机科学与技术学院,吉林,长春,130012
2. 吉林大学符号计算与知识工程教育部重点实验室,吉林,长春,130012
3. 吉林大学公共计算机教学与研究中心,吉林,长春,130012
4. 吉林大学计算机科学与技术学院,吉林,长春,130012
5. 吉林大学符号计算与知识工程教育部重点实验室,吉林,长春,130012
6. 吉林大学公共计算机教学与研究中心,吉林,长春,130012
网络出版:2016-11-25,
纸质出版:2016
移动端阅览
吕颖达, 申铉京, 陈海鹏. 具有几何不变性的图像复制-粘贴盲鉴别算法[J]. 电子学报, 2016,44(11):2592-2599.
L, Uuml, Ying-da. Blind Forensic for Image Copy-Paste Tampering with Geometric Invariance[J]. Acta Electronica Sinica, 2016, 44(11): 2592-2599.
吕颖达, 申铉京, 陈海鹏. 具有几何不变性的图像复制-粘贴盲鉴别算法[J]. 电子学报, 2016,44(11):2592-2599. DOI: 10.3969/j.issn.0372-2112.2016.11.005.
L, Uuml, Ying-da. Blind Forensic for Image Copy-Paste Tampering with Geometric Invariance[J]. Acta Electronica Sinica, 2016, 44(11): 2592-2599. DOI: 10.3969/j.issn.0372-2112.2016.11.005.
提出一种具有几何不变性的复制-粘贴盲鉴别算法.该算法首先定义灰度级别和灰度结构,根据像素的灰度级别划分图像块,再结合其灰度结构定位可疑块对;然后对每一组可疑块对进行基于对数极坐标变换的相位相关,并粗略定位篡改区域;最后,精确计算篡改区域之间的旋转角度、缩放尺度及平移量,从而定位复制-粘贴区域.实验结果表明,该算法不但具有旋转、缩放等几何不变性,而且具有较高检测效率及较强的鲁棒性.
A blind identification algorithm for image copy-paste tampering with geometric invariance was proposed in this study.Firstly
gray level and gray structure are defined
the image is divided into blocks according to the gray levels of pixels
and the suspicious blocks in the image is positioned through the gray structure;Secondly
phase correlation is made for each group of suspicious block based on logarithm polar coordinate transformation
and the tampering regions are located preliminarily;Finally
the rotation angle
scale and translation measurements are calculated precisely
and the tampering regions are positioned.The experimental results show that our algorithm performs well against copy-rotate/scale-paste tampering
and its detection efficiency and robustness are better than the existing algorithms.
0
浏览量
861
下载量
4
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621