基于对抗颜色空间梯度SURF特征匹配复制黏贴篡改检测
|
摘要
为了提高篡改检测方法对复制区域几何变换的鲁棒性,以及解决常规匹配算法带来的误匹配问题,提出了一种基于对抗颜色空间梯度SURF特征匹配复制黏贴篡改检测方法,首先,将完成图像对抗颜色空间梯度的获取,采用梯度值替代灰度值完成SURF的128维特征描述子提取。然后,采用凝聚层次聚类算法对匹配特征关键点进行聚类处理,采用仿射变换估计对聚类后匹配关键点进行估计。最后,采用RANSAC算法优化估计结果。通过实验验证,该方法能准确实现特征点的精确匹配,并能准确定位复制粘贴区域,且对篡改区域的几何变换具有很强的鲁棒性。
In order to improve the robustness of the detection method to the geometric transformation of the replica area, and solve the problem of mismatch caused by the conventional matching algorithm, this paper proposes a space based on the counter color gradient SURF feature matching copy paste tamper detection method. First of all, it will complete the gradient image against color space to obtain, using gradient value instead of grey value 128 D SURF feature descriptor extraction. Then, the aggregation hierarchical clustering algorithm is used to cluster the key points of matching features, and affine transformation estimation is used to estimate the key points of matching after clustering. Finally, RANSAC algorithm is adopted to optimize the estimation results. Experimental results show that the proposed method can match feature points and locate copy-move area accurately, it has strong robustness to the geometric transformation of the tampered area.
In order to improve the robustness of the detection method to the geometric transformation of the replica area, and solve the problem of mismatch caused by the conventional matching algorithm, this paper proposes a space based on the counter color gradient SURF feature matching copy paste tamper detection method. First of all, it will complete the gradient image against color space to obtain, using gradient value instead of grey value 128 D SURF feature descriptor extraction. Then, the aggregation hierarchical clustering algorithm is used to cluster the key points of matching features, and affine transformation estimation is used to estimate the key points of matching after clustering. Finally, RANSAC algorithm is adopted to optimize the estimation results. Experimental results show that the proposed method can match feature points and locate copy-move area accurately, it has strong robustness to the geometric transformation of the tampered area.
引文
[1] BAYRAM S,SENCAR H T,MEMON N.A survey of copy-move forgery detection techniques [C].IEEE Western New York Image Processing Workshop,2008.
[2] LI Y,LIU N, ZHANG B. FI-SURF algorithm for image copy-flip-move forgery detection[J].Journal on Communications,2015,36(5):1-12.
[3] FRIDRICH A J,SOUKAL B D,LUKAS J. Detection of copy-move forgery in digital images [C].Digital Forensic Research Workshop, 2003.
[4] HUANG Y,LU W,SUN W, et al. Improved DCT-based detection of copy-move forgery in image [J].Forensic Science International,2011,206(1/2/3):178-184.
[5] CHRISTLEIN V,RIESS C, ANGELOPOULOU E. On rotation in-variance in copy-move forgery detection [C].Proceeding of the Information Forensics and Security(WIFS),2010:1-6.
[6] AMERINI I,BALLAN L,CALDELLI R, et al. A SIFT-based forensic method for copy-move attack detection and transformation recovery [J]. IEEE Transactions on Information forensics & Security, 2011, 6(3): 1099-1109.
[7] HUANG H,GUO W,ZHANG Y. Detection of copy-move forgery in digital images using sift algorithm [C]. IEEE Pacific-Asia Workshop on Computational Intelligence and Industrial Application, IEEE Computer Society,2008:272-276.
[8] XU B,WANG J,LIU G. Image copy-move forgery detection based on SURF [C].Proceedings of the International Conference on Multimedia Information Networking and Security(MINES),2010:889-892.
[9] 金媛媛.基于SURF的copy-move篡改检测方法研究[D].西安:西北大学,2011.JIN Y Y.SPectral-clustering-based SURF for copy-move attacks[D]. Xi’an: Northwest University,2011.
[10] LIN J,HUANG T Q, LIN L P. Detection algorithm for image copy-move forgery using local intensity order pattern[J]. Journal on Communications,2016,37(Z1):132-139.
[11] 黄跃凯,徐丹,曾昊,等.基于颜色对抗和旋转对称的路标检测算法[J].电子设计工程,2014,22(19):169-172.HUANG Y K, XU D, ZENG H,et al.Road sign detection based on opponent color and rotational symmetry[J].Electronic Design Engineering, 2014,22(19):169-172.
[12] 李春忠,徐宗本,乔琛.带信息反馈的凝聚层次聚类算法[J].中国科学:信息科学,2012,42(6):730-742.LI CH ZH, XU Z B,QIAO CH.Hierarchical agglomerative clustering with information feedback[J]. Science China:Information Sciences, 2012,42(6):730-742.
[13] 谷宗运,谭红春,殷云霞,等.基于SURF和改进的RANSAC算法的医学图像配准[J].中国医学影像学杂志,2014,22(6):470-475,480.GU Z Y, TAN H CH, YIN Y X,et al.Medical Image Registration Based on SURF and Improved RANSAC Algorithm[J]. Chinese Journal of Medical Imaging, 2014,22(6):470-475,480.
[14] BAY H,ESSA A,TUYTELAARS T. Speed-up robust feature [J].Computer Vision and Image Understanding,2008,110(3):346-359.
[15] CASIA V2.0.CASIA V2.0 tampered image detection evaluation(TIDE) database, v2.0(2011) [EB/OL]. http://fore-nsics.idealtest.org.
[2] LI Y,LIU N, ZHANG B. FI-SURF algorithm for image copy-flip-move forgery detection[J].Journal on Communications,2015,36(5):1-12.
[3] FRIDRICH A J,SOUKAL B D,LUKAS J. Detection of copy-move forgery in digital images [C].Digital Forensic Research Workshop, 2003.
[4] HUANG Y,LU W,SUN W, et al. Improved DCT-based detection of copy-move forgery in image [J].Forensic Science International,2011,206(1/2/3):178-184.
[5] CHRISTLEIN V,RIESS C, ANGELOPOULOU E. On rotation in-variance in copy-move forgery detection [C].Proceeding of the Information Forensics and Security(WIFS),2010:1-6.
[6] AMERINI I,BALLAN L,CALDELLI R, et al. A SIFT-based forensic method for copy-move attack detection and transformation recovery [J]. IEEE Transactions on Information forensics & Security, 2011, 6(3): 1099-1109.
[7] HUANG H,GUO W,ZHANG Y. Detection of copy-move forgery in digital images using sift algorithm [C]. IEEE Pacific-Asia Workshop on Computational Intelligence and Industrial Application, IEEE Computer Society,2008:272-276.
[8] XU B,WANG J,LIU G. Image copy-move forgery detection based on SURF [C].Proceedings of the International Conference on Multimedia Information Networking and Security(MINES),2010:889-892.
[9] 金媛媛.基于SURF的copy-move篡改检测方法研究[D].西安:西北大学,2011.JIN Y Y.SPectral-clustering-based SURF for copy-move attacks[D]. Xi’an: Northwest University,2011.
[10] LIN J,HUANG T Q, LIN L P. Detection algorithm for image copy-move forgery using local intensity order pattern[J]. Journal on Communications,2016,37(Z1):132-139.
[11] 黄跃凯,徐丹,曾昊,等.基于颜色对抗和旋转对称的路标检测算法[J].电子设计工程,2014,22(19):169-172.HUANG Y K, XU D, ZENG H,et al.Road sign detection based on opponent color and rotational symmetry[J].Electronic Design Engineering, 2014,22(19):169-172.
[12] 李春忠,徐宗本,乔琛.带信息反馈的凝聚层次聚类算法[J].中国科学:信息科学,2012,42(6):730-742.LI CH ZH, XU Z B,QIAO CH.Hierarchical agglomerative clustering with information feedback[J]. Science China:Information Sciences, 2012,42(6):730-742.
[13] 谷宗运,谭红春,殷云霞,等.基于SURF和改进的RANSAC算法的医学图像配准[J].中国医学影像学杂志,2014,22(6):470-475,480.GU Z Y, TAN H CH, YIN Y X,et al.Medical Image Registration Based on SURF and Improved RANSAC Algorithm[J]. Chinese Journal of Medical Imaging, 2014,22(6):470-475,480.
[14] BAY H,ESSA A,TUYTELAARS T. Speed-up robust feature [J].Computer Vision and Image Understanding,2008,110(3):346-359.
[15] CASIA V2.0.CASIA V2.0 tampered image detection evaluation(TIDE) database, v2.0(2011) [EB/OL]. http://fore-nsics.idealtest.org.