Blind identification of image copy-move forgery based on feature points and dense clustering

被引：0
|
作者：
Chen H.-P. [1 ,2 ]
Qu Z.-D. [2 ,3 ]
Yang X.-W. [2 ,3 ]
Zhang W. [4 ]
Lyu Y.-D. [5 ]
机构：
[1] College of Computer Science and Technology, Jilin University, Changchun
[2] Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Changchun
[3] College of Software, Jilin University, Changchun
[4] The Second Hospital, Jilin University, Changchun
[5] Public Computer Education and Research Center, Jilin University, Changchun
来源：
Jilin Daxue Xuebao (Gongxueban)/Journal of Jilin University (Engineering and Technology Edition) | 2020年 / 50卷 / 03期
关键词：
AKAZE feature; Copy-move forgery detection; Copy-move forgery localization; Dense clustering; Digital image blind forensics; Information processing technology;
D O I：
10.13229/j.cnki.jdxbgxb20190517
中图分类号：
学科分类号：
摘要：
Image manipulation forensics technology has become a hot research topic in the field of image processing. Copy-move forgery detection is one of the most important and popular digital image forensics technologies. In order to solve the problems of high time complexity and inaccurate location of tampered regions in digital image copy-move forgery, a method of copy-paste forgery detection based on binary descriptor and density clustering algorithm is proposed. First, the binary descriptor AKAZE is used to extract feature points and describe features to reduce time complexity of feature calculation. Second, the density-based clustering algorithm DBSCAN is applied to remove false matches. Finally, the tampered area is located by affine transformation of the whole image. The experimental results on different datasets show that using the proposed algorithm, the highest localization accuracy exceeds 95%, and the false detection rate is basically less than 10%. It can effectively reduce the time complexity and show better performance in reliability and accuracy of forgery location. © 2020, Jilin University Press. All right reserved.
引用
收藏
页码：1069 / 1076
页数：7
相关论文
共 13 条
  • [1] Lu Y.-D., Shen X.-J., Chen H.-P., Blind forensic for image copy-paste tampering with geometric invariance, Acta Electronica Sinica, 44, 11, pp. 2592-2599, (2016)
  • [2] Li Y., Image copy-move forgery detection based on polar cosine transform and approximate nearest neighbor searching, Forensic Science International, 224, 1-3, pp. 59-67, (2013)
  • [3] Shen X.-J., Zhu Y., Lyu Y.-D., Et al., Coloured image copy-move forgery detection based on SIFT and HSI, Journal of Jilin University(Engineering and Technology Edition), 44, 1, pp. 171-176, (2014)
  • [4] Chen Y.-H., Zhang D.-X., Yang S.-S., Et al., Image multiple copy-move forgery detection based on SURF algorithm, Computer Engineering and Design, 39, 8, pp. 2593-2597, (2018)
  • [5] Ardizzone E., Bruno A., Mazzola G., Copy-move forgery detection by matching triangles of keypoints, IEEE Transactions on Information Forensics and Security, 10, 10, pp. 2084-2094, (2015)
  • [6] Pun C.M., Yuan X.C., Bi X.L., Image forgery detection using adaptive oversegmentation and feature point matching, IEEE Transactions on Information Forensics and Security, 10, 8, pp. 1705-1716, (2015)
  • [7] Sun X., Liu Z., Hu Y., Et al., Perceptual multi-channel visual feature fusion for scene categorization, Information Sciences, 429, pp. 37-48, (2018)
  • [8] Xia Y., Zhang L., Liu Z., Et al., Weakly-supervised multimodal kernel for categorizing aerial photographs, IEEE Trans Image Process, 26, 8, pp. 3748-3758, (2017)
  • [9] Wu H.-Q., Li C.-C., Xie Y., Application of improved A-KAZE algorithm in image registration, Journal of Southeast University(Natural Science Edition), 47, 4, pp. 668-672, (2017)
  • [10] Yang X., Cheng K.T., LDB: an ultra-fast feature for scalable augmented reality on mobile devices, 2012 IEEE International Symposium on Mixed and Augmented Reality, pp. 49-57, (2012)
12