基于深度特征提取和DCT变换的图像复制粘贴篡改检测

计算机工程与科学 ›› 2023, Vol. 45 ›› Issue (01): 163-170.

基于深度特征提取和DCT变换的图像复制粘贴篡改检测

魏伟一,赵毅凡,陈帼

(西北师范大学计算机科学与工程学院，甘肃兰州 730070)

收稿日期:2021-05-21 修回日期:2021-10-08 接受日期:2023-01-25 出版日期:2023-01-25 发布日期:2023-01-25
基金资助:
甘肃省自然科学基金（20JR5RA518）

Image copy-move forgery detection based on deep feature extraction and DCT transform

WEI Wei-yi,ZHAO Yi-fan,CHEN Guo

(College of Computer Science and Engineering,Northwest Normal University,Lanzhou 730070,China)

Received:2021-05-21 Revised:2021-10-08 Accepted:2023-01-25 Online:2023-01-25 Published:2023-01-25

摘要/Abstract

摘要： 针对图像复制粘贴篡改检测中算法时间复杂度过高和定位区域不完整的问题，提出一种基于深度特征提取和离散余弦变换的图像复制粘贴篡改检测算法。首先，融合图像颜色和纹理信息获得四通道图像，计算自适应特征提取阈值，并通过基于全卷积神经网络的特征检测器提取图像深度特征；其次，通过离散余弦变换提取块特征进行初步匹配，再利用点特征向量消除误匹配；最后，通过卷积运算精确定位篡改区域。通过在公共数据集上进行验证，充分展示了该算法在检测效率和定位区域完整性方面的优势。

关键词: 复制粘贴篡改, 深度特征, 离散余弦变换, 卷积运算

Abstract: Aiming at the problems of high algorithm time complexity and incomplete location region in image copy-move forgery detection, an image copy-move forgery detection algorithm based on deep features extraction and discrete cosine transform is proposed. Firstly, the color and texture information of the image are fused to obtain the four-channel image, the adaptive feature extraction threshold is calculated, and the depth feature is extracted by the feature detector based on the full convolutional neural network. Secondly, the discrete cosine transform is used to extract block features for preliminary matching, and point feature vectors are adopted to eliminate mismatches. Finally, the tampered regions are accurately located by convolution operation. The verification on the public dataset fully demonstrates the advantages of the algorithm in the detection efficiency and the integrity of the positioning area.

Key words: copy-move forgery detection, deep feature, discrete cosine transform, convolution operation

魏伟一, 赵毅凡, 陈帼. 基于深度特征提取和DCT变换的图像复制粘贴篡改检测[J]. 计算机工程与科学, 2023, 45(01): 163-170.

WEI Wei-yi, ZHAO Yi-fan, CHEN Guo. Image copy-move forgery detection based on deep feature extraction and DCT transform[J]. Computer Engineering & Science, 2023, 45(01): 163-170.

参考文献［8］

［1］	Ikhlayel M,Hariadi M,Pumama I.Modified multi-scale feature extraction for copy-move forgery detection based on CMD-SIFT［C］∥Proc of 2018 International Conference on Computer Engineering,Network and Intelligent Multimedia,2018:260-264.
［2］	Alhammadi M M, Emmanuel S. Improving SURF based copy-move forgery detection using super resolution［C］∥Proc of IEEE International Symposium on Multimedia,2016:341-344.
［3］	Donoser M, Schmalstieg D.Discriminative feature-to-point matching in image-based localization［C］∥Proc of 2014 IEEE Conference on Computer Vision and Pattern Recognition,2014:516-523.
［4］	Yang H Y,Qi S R,Niu Y,et al.Copy-move forgery detection based on adaptive keypoints extraction and matching［J］.Multimedia Tools and Applications,2019,78:34585-34612.
［5］	Niyishaka P,Bhagvati C.Copy-move forgery detection using image blobs and BRISK feature［J］.Multimedia Tools and Applications,2020,79:26045-26059.
［6］	Li Y,Zhou J.Fast and effective image copy-move forgery detection via hierarchical feature point matching［J］.IEEE Transactions on Information Forensics and Security,2019,14(5):1307-1322.
［7］	Lin C,Lu W,Huang X C,et al.Copy-move forgery detection using combined features and transitive matching［J］.Multimedia Tools and Applications,2019,78:30081-30096.
［8］	Liu An-an, Wang Hong-tao.A blind copy-move forensics method based on truncated DCT domain coefficients:CN201710365547.5［P］.2017-09-26.(in Chinese)
［9］	Jwaid M F,Baraskar T N.Detection of copy-move image forgery using local binary pattern with discrete wavelet transform and principle component analysis［C］∥Proc of International Conference on Computing,Communication,Control and Automation,2017:1-6.
［10］	Pun C M, Chuang J L. A two-stage localization for copy-move forgery detection［J］.Information Sciences,2018,463-464:33-35.
［11］	Chen C C,Lu W Y,Chou C H.Rotational copy-move forgery detection using SIFT and region growing strategies［J］.Multimedia Tools and Applications,2019,78:18293-18308.
［12］	Muja M, Lowe D G. Scalable nearest neighbor algorithms for high dimensional data［J］. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2014, 36(11):2227-2240.
［13］	Ryu S J,Kirchner M,Lee M J,et al.Rotation invariant localization of duplicated image regions based on Zernike moments［J］.IEEE Transactions on Information Forensics & Security,2013,8(8):1355-1370.
［14］	Barnes C,Shechtman E,Goldman D B,et al.The generalized PatchMatch correspondence algorithm［C］∥Proc of European Conference on Computer Vision,2010:29-43.
［15］	Chen B J, Gao Y, Yu M, et al.An effective copy-move forgery detection algorithm using fractional quaternion Zernike moments and improved PatchMatch algorithm［J］.Journal of Southeast University(English Edition),2019,35(4):431-439.
［16］	Bi X L,Pun C M,Yuan X C.Multi-scale feature extraction and adaptive matching for copy-move forgery detection［J］.Multimedia Tools and Applications,2018,77:363-385.
［17］	Liu Y Q,Guan Q X,Zhao X F.Copy-move forgery detection based on convolutional kernel network［J］.Multimedia Tools and Applications,2018,77:18269-18293.
［18］	Agarwal R,Verma O P.An efficient copy move forgery detection using deep learning feature extraction and matching algorithm［J］.Multimedia Tools and Applications,2020,79:1-22.
［19］	Cozzolino D, Poggi G, Verdoliva L. Efficient dense-field copy-move forgery detection［J］.IEEE Transactions on Information Forensics and Security,2015,10(11):2284-2297.
［20］	DeTone D,Malisiewicz T,Rabinovich A.SuperPoint:Self-supervised interest point detection and description［C］∥Proc of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops,2018:337-349.
［21］	Tralic D,Zupancic I,Grgic S,et al.CoMoFoD—New database for copy-move forgery detection［C］∥Proc of Croatian Society Electronics in Marine International Symposium,2013:49-54.
	附中文参考文献：
［8］	刘安安,王洪涛.一种基于截断DCT域系数的复制粘贴盲取证方法:CN201710365547.5［P］.2017-09-26.

[1]	周燕, 潘丽丽, 陈蓉玉, 邵伟志, 雷前慧. 结合自适应融合网络与哈希的图像检索算法 [J]. 计算机工程与科学, 2021, 43(09): 1616-1622.
[2]	王若飞, 刘锋. 选择排序的DCT系数对JPEG图像的可逆数据隐藏[J]. 计算机工程与科学, 2020, 42(11): 2013-2019.
[3]	谭志国1，2,欧建平1,张军1,沈先耿2. 多特征组合的深度图像分割算法[J]. 计算机工程与科学, 2018, 40(08): 1429-1434.
[4]	熊祥光，韦立，谢刚. 基于3D-DCT和SVD的鲁棒彩色图像水印算法[J]. J4, 2015, 37(06): 1093-1100.
[5]	熊祥光，王端理. 基于HVS和关系的DCT域彩色图像水印方案[J]. J4, 2014, 36(02): 311-316.
[6]	谢志华1,孙俊杰2. 基于简化血流图小波包域DCT系数决策融合的红外人脸识别[J]. J4, 2011, 33(9): 184-188.
[7]	祖研1，帅仁俊1，陈平2. 基于DICOM标准的医学图像有损压缩的研究[J]. J4, 2011, 33(7): 62-66.
[8]	刘凌霞1，牛红惠1，崔洲涓2. 基于分块DCT的视频文字检测算法[J]. J4, 2011, 33(6): 63-66.
[9]	高一佳，李陶深. NURBS自由曲线曲面数据优化压缩方法[J]. J4, 2010, 32(9): 43-46.
[10]	赵永耀秦志光. 图像通信中去块效应方法研究现状[J]. J4, 2008, 30(5): 37-40.
[11]	韩金姝. 结合DCT变换的快速分形图像编码方法[J]. J4, 2007, 29(8): 52-54.
[12]	陈哲张永林. 数字水印技术在二维条码证件防伪中的应用[J]. J4, 2006, 28(4): 42-44.
[13]	毛万欣卢正鼎. 一种基于DCT的自适应图像数字水印算法[J]. J4, 2005, 27(7): 43-45.
[14]	蒋加伏李广琼唐贤瑛. 基于小波包分析的ATM网分层传输 [J]. J4, 2004, 26(6): 47-49.