多距离特征匹配的篡改图像检测算法
|
摘要
为了解决当前篡改图像的检测算法主要依靠单一的特征进行描述以及欧几里德距离进行匹配,导致篡改图像的检测率较低的问题,以及在对图像复制粘贴后进行一系列后处理操作的篡改图像检测时,容易出现匹配错误和鲁棒性差的问题,采用一种多距离特征匹配的篡改图像检测算法。首先,对获取到的图像提取尺度不变特征变换(SIFT)特征,在SIFT特征待描述区域的基础上,提取具有权值旋转不变均匀性的局部二值模式(LBP)特征,构成特征描述子;其次,分别计算描述子之间的标准欧几里德距离、相关距离以及汉明距离,通过多距离匹配改进g2nn算法进行特征的初次匹配;最后,通过凝聚型分层特征聚类以及随机一致性(RANSAC)算法去除存在的错误匹配点,完成篡改图像的检测。在MICC-F220图像数据库上进行了测试,实验结果表明,与当前2种主流算法相比,总体准确率分别提高了2.86%和2.11%,对于缩放、旋转以及缩放+旋转的后处理均具有很好的鲁棒性,是一种研究复制粘贴后进行缩放和旋转后处理的篡改图像检测的有效方法。
In order to solve the problem that the detection algorithm of the current tampering image relies mainly on a single feature to describe and the Euclidean distance to match, the detection rate of the tampering image is comparatively low, and matching errors and poor robustness are prone to occur when a series of post processing tampering images are detected after copy paste images, a multi-distance feature matching detection algorithm is used in this paper. Firstly, the scale-invariant feature transform(SIFT) feature was extracted from the acquired image, and the local binary patterns(LBP) feature with the rotation invariance uniformity of the weight was extracted on the basis of the SIFT region to be described, so the feature descriptor was constructed. Secondly, the standard Euclidean distance, correlation distance and hamming distance were calculated respectively, and G2 NN algorithm was improved by multi-distance matching to perform the initial match of the feature. Finally, mismatched points were removed by condensed hierarchical feature clustering and random sample consensus(RANSAC) algorithm to complete tampering image detection.The test was carried out on the MICC-F220 image database with the result that the overall accuracy of the proposed algorithm is improved by 2.86% and 2.11% respectively compared with the two mainstream algorithms available. It is robust to scaling, rotation and scaling and rotation post-processing. It is an effective method to detect tampering image detection after copying and pasting and then performing scaling and rotation processing.
In order to solve the problem that the detection algorithm of the current tampering image relies mainly on a single feature to describe and the Euclidean distance to match, the detection rate of the tampering image is comparatively low, and matching errors and poor robustness are prone to occur when a series of post processing tampering images are detected after copy paste images, a multi-distance feature matching detection algorithm is used in this paper. Firstly, the scale-invariant feature transform(SIFT) feature was extracted from the acquired image, and the local binary patterns(LBP) feature with the rotation invariance uniformity of the weight was extracted on the basis of the SIFT region to be described, so the feature descriptor was constructed. Secondly, the standard Euclidean distance, correlation distance and hamming distance were calculated respectively, and G2 NN algorithm was improved by multi-distance matching to perform the initial match of the feature. Finally, mismatched points were removed by condensed hierarchical feature clustering and random sample consensus(RANSAC) algorithm to complete tampering image detection.The test was carried out on the MICC-F220 image database with the result that the overall accuracy of the proposed algorithm is improved by 2.86% and 2.11% respectively compared with the two mainstream algorithms available. It is robust to scaling, rotation and scaling and rotation post-processing. It is an effective method to detect tampering image detection after copying and pasting and then performing scaling and rotation processing.
引文
[1]Korus P,Huang J.Multi-scale fusion for improved localization of malicious tampering in digital images[J].IEEETransactions on Image Processing,2016,25(3):1312-1326.
[2]赵超然.基于超像素分割的图像拼接定位检测算法研究[D].长春:吉林大学,2018.ZHAO Chao-ran.Research on image splicing localization algorithm based on super pixel segmentation[D].Changchun:Jilin University,2018.
[3]周学花.基于特征点的复制-粘贴篡改图像盲鉴别与定位算法研究[D].长春:吉林大学,2018.ZHOU Xue-hua.Research on blind identification and localization algorithm of copy-move forgery image based on key-point[D].Changchun:Jilin University,2018.
[4]彭小洋.基于块特征向量匹配的图像复制-粘贴被动取证算法研究[D].成都:西南交通大学,2018.PENG Xiao-yang.Copy-move image forensics algorithm based on block feature vectors matching[D].Chengdu:Southwest Jiaotong University,2018.
[5]Lowe D G.Distinctive image features from scale-invariant key points[J].International Journal of Computer Vision,2004,60(2):91-110.
[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 and Security,2011,6(3):1099-1110.
[7]Bay H,Ess A,Tuytelaars T,et al.SURF:speeded up robust features[J].Computer Vision and Image Understanding,2008,110(3):346-359.
[8]Zhu Y,Ng T,Wen B,et al.Copy-move forgery detection in the presence of similar but genuine objects[C]//2017 IEEE 2nd International Conference on Signal and Image Processing(ICSIP),Singapore,2017:25-29.
[9]陈敏,汤晓安.SIFT与SURF特征提取算法在图像匹配中的应用对比研究[J].现代电子技术,2018,41(7):41-44.CHEN Min,TANG Xiao-an.Comparison study on application of SIFT and SURF feature extraction algorithms in image matching[J].Modern Electronics Technique,2018,41(7):41-44.
[10]杜振龙,杨凡,李晓丽,等.利用SIFT特征的非对称匹配图像拼接盲检测[J].中国图象图形学报,2013,18(4):442-449.DU Zhen-long,YANG Fan,LI Xiao-li,et al.Forgery image blind detection by asymmetric search based on SIFT[J].Journal of Image and Graphics,2013,18(4):442-449.
[11]李岩,刘念,张斌,等.图像镜像复制粘贴篡改检测中的FI-SURF算法[J].通信学报,2015,36(5):58-69.LI Yan,LIU Nian,ZHANG Bin,et al.FI-SURF algorithm for image copy-flip-move forgery detection[J].Journal on Communications,2015,36(5):58-69.
[12]李昆仑,孙硕.基于改进SIFT算法的图像复制粘贴篡改检测[J].计算机科学,2016,43(6A):179-183.LI Kun-lun,SUN Shuo.Image copy-paste tampering detection based on improved SIFT algorithm[J].Computer Science,2016,43(6A):179-183.
[13]Swapnil H K,Avinash D G.Copy-Move attack forgery detection by using SIFT[J].International Journal of Innovative Technology and Exploring Engineering,2013,2(5):221-224.
[14]Shahroudnejad A,Rahmati M.Copy-move forgery detection in digital images using affine-SIFT[C]//2016 2nd International Conference of Signal Processing and Intelligent Systems(ICSPIS),Tehran,2016:1-5.
[15]柴建伟,刘婷.改进的SIFT耦合特征点集群的图像伪造检测算法[J].西南师范大学学报(自然科学版),2018,43(3):34-41.CHAI Jian-wei,LIU Ting.Image forgery detection algorithm based on improved SIFT coupled feature point clustering[J].Journal of Southwest China Normal University(Natural Science Edition),2018,43(3):34-41.
[16]刘丽,谢毓湘,魏迎梅,等.局部二进制模式方法综述[J].中国图象图形学报,2014,19(12):1696-1720.LIU Li,XIE Yu-xiang,WEI Ying-mei,et al.Survey of local binary pattern method[J].Journal of Image and Graphics,2014,19(12):1696-1720.
[17]邓少闻,罗代升,郭崇.多尺度LBP耦合K-D树的图像伪造盲检测算法[J].计算机工程与设计,2017,38(5):1307-1313.DENG Shao-wen,LUO Dai-sheng,GUO Chong.Image forgery blind detection algorithm based on multi-scale LBP coupled K-D tree[J].Computer Engineering and Design,2017,38(5):1307-1313.
[18]郭继昌,王秋子,赵洁,等.一种基于LBP和马尔科夫特征的细缝裁剪取证方法[J].电子科技大学学报,2018,47(4):481-485.GUO Ji-chang,WANG Qiu-zi,ZHAO Jie,et al.A method of seam carving forensics based on LBP and Markov features[J].Journal of University of Electronic Science and Technology,2018,47(4):481-485.
[19]郑永斌,黄新生,丰松江.SIFT和旋转不变LBP相结合的图像匹配算法[J].计算机辅助设计与图形学学报,2010,22(2):286-292.ZHENG Yong-bin,HUANG Xin-sheng,FENG Songjiang.An image matching algorithm based on combination of SIFT and the rotation invariant LBP[J].Journal of Computer-Aided Design&Computer Graphics,2010,22(2):286-292.
[20]邵虹,朱虹,崔文成.抗翻转、旋转和缩放攻击的图像区域复制篡改检测[J].计算机辅助设计与图形学学报,2015,27(1):157-165.SHAO Hong,ZHU Hong,CUI Wen-cheng.Detection of image region-duplication forgery affected by flipping,rotation and scaling[J].Journal of Computer-Aided Design&Computer Graphics,2015,27(1):157-165.
[21]陈辉映,张大兴,杨珊珊,等.基于SURF的图像多区域复制粘贴篡改检测[J].计算机工程与设计,2018,39(8):2593-2597.CHEN Hui-ying,ZHANG Da-xing,YANG Shan-shan,et al.Image multiple copy-move forgery detection based on SURF algorithm[J].Computer Engineering and Design,2018,39(8):2593-2597.
[22]甘玲,周灿,李大港.一种采用点匹配的图像区域复制粘贴篡改检测方法[J].小型微型计算机系统,2017,38(7):1631-1635.GAN Ling,ZHOU Can,LI Da-gang.Method of digital image region copy-move forgery detection using point matching[J].Journal of Chinese Computer Systems,2017,38(7):1631-1635.
[23]杨雨薇,张亚萍.一种改进的SIFT图像检测与特征匹配算法[J].云南大学学报(自然科学版),2017,39(3):376-384.YANG Yu-wei,ZHANG Ya-ping.An improved SIFT image detection and feature matching algorithm[J].Journal of Yunnan University(Natural Science Edition),2017,39(3):376-384.
[24]Li J,Li X,Yang B,et al.Segmentation based image copy-move forgery detection scheme[J].IEEE Transactions on Information Forensics and Security,2015,10(3):507-518.
[25]汪磊,曾宪庭,苏金阳.一种基于多域特征的JPEG图像隐写分析算法[J].计算机科学,2014,41(6):94-98.WANG Lei,ZENG Xian-ting,SU Jin-Yang.Steg analysis based on multi-domain features for JPEG images[J].Computer Science.2014,41(6):94-98.
[2]赵超然.基于超像素分割的图像拼接定位检测算法研究[D].长春:吉林大学,2018.ZHAO Chao-ran.Research on image splicing localization algorithm based on super pixel segmentation[D].Changchun:Jilin University,2018.
[3]周学花.基于特征点的复制-粘贴篡改图像盲鉴别与定位算法研究[D].长春:吉林大学,2018.ZHOU Xue-hua.Research on blind identification and localization algorithm of copy-move forgery image based on key-point[D].Changchun:Jilin University,2018.
[4]彭小洋.基于块特征向量匹配的图像复制-粘贴被动取证算法研究[D].成都:西南交通大学,2018.PENG Xiao-yang.Copy-move image forensics algorithm based on block feature vectors matching[D].Chengdu:Southwest Jiaotong University,2018.
[5]Lowe D G.Distinctive image features from scale-invariant key points[J].International Journal of Computer Vision,2004,60(2):91-110.
[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 and Security,2011,6(3):1099-1110.
[7]Bay H,Ess A,Tuytelaars T,et al.SURF:speeded up robust features[J].Computer Vision and Image Understanding,2008,110(3):346-359.
[8]Zhu Y,Ng T,Wen B,et al.Copy-move forgery detection in the presence of similar but genuine objects[C]//2017 IEEE 2nd International Conference on Signal and Image Processing(ICSIP),Singapore,2017:25-29.
[9]陈敏,汤晓安.SIFT与SURF特征提取算法在图像匹配中的应用对比研究[J].现代电子技术,2018,41(7):41-44.CHEN Min,TANG Xiao-an.Comparison study on application of SIFT and SURF feature extraction algorithms in image matching[J].Modern Electronics Technique,2018,41(7):41-44.
[10]杜振龙,杨凡,李晓丽,等.利用SIFT特征的非对称匹配图像拼接盲检测[J].中国图象图形学报,2013,18(4):442-449.DU Zhen-long,YANG Fan,LI Xiao-li,et al.Forgery image blind detection by asymmetric search based on SIFT[J].Journal of Image and Graphics,2013,18(4):442-449.
[11]李岩,刘念,张斌,等.图像镜像复制粘贴篡改检测中的FI-SURF算法[J].通信学报,2015,36(5):58-69.LI Yan,LIU Nian,ZHANG Bin,et al.FI-SURF algorithm for image copy-flip-move forgery detection[J].Journal on Communications,2015,36(5):58-69.
[12]李昆仑,孙硕.基于改进SIFT算法的图像复制粘贴篡改检测[J].计算机科学,2016,43(6A):179-183.LI Kun-lun,SUN Shuo.Image copy-paste tampering detection based on improved SIFT algorithm[J].Computer Science,2016,43(6A):179-183.
[13]Swapnil H K,Avinash D G.Copy-Move attack forgery detection by using SIFT[J].International Journal of Innovative Technology and Exploring Engineering,2013,2(5):221-224.
[14]Shahroudnejad A,Rahmati M.Copy-move forgery detection in digital images using affine-SIFT[C]//2016 2nd International Conference of Signal Processing and Intelligent Systems(ICSPIS),Tehran,2016:1-5.
[15]柴建伟,刘婷.改进的SIFT耦合特征点集群的图像伪造检测算法[J].西南师范大学学报(自然科学版),2018,43(3):34-41.CHAI Jian-wei,LIU Ting.Image forgery detection algorithm based on improved SIFT coupled feature point clustering[J].Journal of Southwest China Normal University(Natural Science Edition),2018,43(3):34-41.
[16]刘丽,谢毓湘,魏迎梅,等.局部二进制模式方法综述[J].中国图象图形学报,2014,19(12):1696-1720.LIU Li,XIE Yu-xiang,WEI Ying-mei,et al.Survey of local binary pattern method[J].Journal of Image and Graphics,2014,19(12):1696-1720.
[17]邓少闻,罗代升,郭崇.多尺度LBP耦合K-D树的图像伪造盲检测算法[J].计算机工程与设计,2017,38(5):1307-1313.DENG Shao-wen,LUO Dai-sheng,GUO Chong.Image forgery blind detection algorithm based on multi-scale LBP coupled K-D tree[J].Computer Engineering and Design,2017,38(5):1307-1313.
[18]郭继昌,王秋子,赵洁,等.一种基于LBP和马尔科夫特征的细缝裁剪取证方法[J].电子科技大学学报,2018,47(4):481-485.GUO Ji-chang,WANG Qiu-zi,ZHAO Jie,et al.A method of seam carving forensics based on LBP and Markov features[J].Journal of University of Electronic Science and Technology,2018,47(4):481-485.
[19]郑永斌,黄新生,丰松江.SIFT和旋转不变LBP相结合的图像匹配算法[J].计算机辅助设计与图形学学报,2010,22(2):286-292.ZHENG Yong-bin,HUANG Xin-sheng,FENG Songjiang.An image matching algorithm based on combination of SIFT and the rotation invariant LBP[J].Journal of Computer-Aided Design&Computer Graphics,2010,22(2):286-292.
[20]邵虹,朱虹,崔文成.抗翻转、旋转和缩放攻击的图像区域复制篡改检测[J].计算机辅助设计与图形学学报,2015,27(1):157-165.SHAO Hong,ZHU Hong,CUI Wen-cheng.Detection of image region-duplication forgery affected by flipping,rotation and scaling[J].Journal of Computer-Aided Design&Computer Graphics,2015,27(1):157-165.
[21]陈辉映,张大兴,杨珊珊,等.基于SURF的图像多区域复制粘贴篡改检测[J].计算机工程与设计,2018,39(8):2593-2597.CHEN Hui-ying,ZHANG Da-xing,YANG Shan-shan,et al.Image multiple copy-move forgery detection based on SURF algorithm[J].Computer Engineering and Design,2018,39(8):2593-2597.
[22]甘玲,周灿,李大港.一种采用点匹配的图像区域复制粘贴篡改检测方法[J].小型微型计算机系统,2017,38(7):1631-1635.GAN Ling,ZHOU Can,LI Da-gang.Method of digital image region copy-move forgery detection using point matching[J].Journal of Chinese Computer Systems,2017,38(7):1631-1635.
[23]杨雨薇,张亚萍.一种改进的SIFT图像检测与特征匹配算法[J].云南大学学报(自然科学版),2017,39(3):376-384.YANG Yu-wei,ZHANG Ya-ping.An improved SIFT image detection and feature matching algorithm[J].Journal of Yunnan University(Natural Science Edition),2017,39(3):376-384.
[24]Li J,Li X,Yang B,et al.Segmentation based image copy-move forgery detection scheme[J].IEEE Transactions on Information Forensics and Security,2015,10(3):507-518.
[25]汪磊,曾宪庭,苏金阳.一种基于多域特征的JPEG图像隐写分析算法[J].计算机科学,2014,41(6):94-98.WANG Lei,ZENG Xian-ting,SU Jin-Yang.Steg analysis based on multi-domain features for JPEG images[J].Computer Science.2014,41(6):94-98.