基于彩色LBP的隐蔽性复制-粘贴篡改盲鉴别算法

朱叶; 申铉京; 陈海鹏

doi:10.16383/j.aas.2017.c160068

基于彩色LBP的隐蔽性复制-粘贴篡改盲鉴别算法

doi: 10.16383/j.aas.2017.c160068

朱叶^1,2,,
申铉京^1,2,,
陈海鹏^1,2, ,

1.
吉林大学计算机科学与技术学院长春 130012
2.
吉林大学符号计算与知识工程教育部重点实验室长春 130012

基金项目:

吉林省自然科学基金 20150101055JC

国家青年科学基金 61305046

吉林省自然科学基金 20140101193JC

国家青年科学基金 61602203

详细信息

作者简介:
朱叶吉林大学计算机科学与技术学院博士研究生.2011年获得山东科技大学学士学位.主要研究方向为图像处理, 数字图像盲取证技术.E-mail:zhuye13@mails.jlu.edu.cn

申铉京吉林大学计算机科学与技术学院教授.1990年获得哈尔滨工业大学博士学位.主要研究方向为多媒体技术, 计算机图像处理, 智能测量系统, 光电混合系统.E-mail:xjshen@jlu.edu.cn

通讯作者:
陈海鹏吉林大学计算机科学与技术学院副教授.主要研究方向为图像处理与模式识别.本文通信作者.E-mail:chenhp@jlu.edu.cn

计量
- 文章访问数: 2367
- HTML全文浏览量: 196
- PDF下载量: 635
- 被引次数: 0
出版历程
- 收稿日期: 2016-01-22
- 录用日期: 2016-04-01
- 刊出日期: 2017-03-20

Covert Copy-move Forgery Detection Based on Color LBP

ZHU Ye^{1,2
,},
SHEN Xuan-Jing^{1,2
,},
CHEN Hai-Peng^{1,2
, ,}

1.
College of Computer Science and Technology, Jilin University, Changchun 130012
2.
Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Jilin University, Changchun 130012

Funds:

Natural Science Foundation of Jilin Province 20150101055JC

the National Science Foundation for Young Scientists of China 61305046

Natural Science Foundation of Jilin Province 20140101193JC

the National Science Foundation for Young Scientists of China 61602203

More Information

Author Bio:
Ph. D. candidate at the College of Computer Science and Technology, Jilin University. She received her bachelor degree from Shandong University of Technology in 2011. Her research interest covers image processing and image forensic technology

Professor at the College of Computer Science and Technology, Jilin University. He received his Ph. D. degree from Harbin Institute of Technology in 1990. His research interest covers multimedia technology, computer image processing, intelligent measurement system, and optical-electronic hybrid system

Corresponding author: CHEN Hai-PengAssociate professor at the College of Computer Science and Technology, Jilin University. His research interest covers image processing and pattern recognition. Corresponding author of this paper

摘要

摘要: 现有的复制-粘贴盲鉴别算法大多忽略图像彩色信息，导致对隐蔽性篡改方式的检测率较低，基于此，本文提出一种基于彩色局部二值模式（Color local binary patterns，CoLBP）的隐蔽性复制-粘贴盲鉴别算法.算法首先对彩色图像进行预处理，即建立彩色LBP纹理图像，从而实现彩色信息与LBP纹理特征的融合；其次重叠分块并提取灰度共生矩阵（Gray level co-occurrence matrix，GLCM）特征；最后，提出改进的kd树和超平面划分标记split搜索方法，快速匹配图像块，并应用形态学操作去除误匹配，精确定位复制-粘贴区域.实验结果表明，本算法对隐蔽性复制-粘贴篡改定位准确，并对模糊、噪声、JPEG重压缩后处理操作有很好的鲁棒性.
- 图像盲鉴别 /
- 复制-粘贴篡改 /
- 彩色局部二值模式 /
- 灰度共生矩阵 /
- kd树
Abstract: Since negligence of color information in detecting copy-move forgeries leads to low accuracy in detection of in covert tampering, a novel method using color local binary patterns (CoLBP) is proposed. It involves the following three steps: first, establish color LBP texture image, which is a preprocessing of image and a combination of color information and LBP texture; second, divide into overlapping blocks and extract gray level co-occurrence matrix (GLCM) features; finally, match image blocks by the improved kd tree and split partition, remove the false matched blocks using morphological operation and then detect the resulting copy-move regions. Experimental results show that our algorithm is effective for covert tampering, and exhibits high robustness even when an image is distorted by blur, noise and JPEG recompression.
- Image blind identification /
- copy-move forgery /
- color local binary patterns (CoLBP) /
- gray level co-occurrence matrix (GLCM) /
- kd tree
注释:

1) 本文责任编委刘成林

HTML全文

图 1 复制-粘贴篡改分类示例 ((a) 隐蔽性复制-粘贴篡改图像示例; (b) 造成假象类复制-粘贴篡改图像示例, 其中实线和虚线矩形框分别为复制和篡改区域)

Fig. 1 The exemplar of copy-move forgery classify ((a) The exemplar of covert copy-move forgery; (b) The exemplar of spurious copy-move forgery, where solid and dashed rectangles are copied and pasted regions.)

下载: 全尺寸图片幻灯片

图 2 本文算法流程图

Fig. 2 The framework of proposed method

下载: 全尺寸图片幻灯片

图 3 彩色LBP图像建立流程图

Fig. 3 The framework of color LBP establishment

下载: 全尺寸图片幻灯片

图 4 GLCM特征提取示意图 (其中实线分别表示o=1~4, s=1;虚线表示o=1, s=2)

Fig. 4 The diagram of GLCM feature extraction, where solid line are respectively indicate o=1~4, s=1; dash line indicates o=1, s=2

下载: 全尺寸图片幻灯片

图 5 不同彩色模型下的彩色LBP图像

Fig. 5 The color LBP images on different color model

下载: 全尺寸图片幻灯片

图 6 彩色LBP图像的有效性实验示例 ((a) 直接在灰度图像上提取GLCM特征匹配结果; (b) 本文算法匹配结果, 其中标记区域为篡改区域)

Fig. 6 The experiment on effectiveness of color LBP ((a) The matching result based on directly extracting GLCM feature on gray image; (b) The matching result based on our method, where marked regions are forged regions.)

下载: 全尺寸图片幻灯片

图 7 不同阈值下的FPR和TRP对比

Fig. 7 The FPR and TPR comparison on different threshold

下载: 全尺寸图片幻灯片

图 8 隐蔽性复制-粘贴篡改检测结果示例 ((a), (g) 隐蔽性复制-粘贴篡改图像; (b), (h) 彩色LBP图像; (c), (i) DCT^[4]算法检测结果; (d), (j) Zernike^[14]算法检测结果; (e), (k) LBP^[17]算法检测结果; (f), (l) 本文算法检测结果.其中, (a), (g) 实线和虚线框分别表示复制和粘贴区域; (c)~(f), (i)~(l) 标记区域为算法检测篡改区域)

Fig. 8 The exemplar results on covert copy-move forgery detection ((a), (g) Covert copy-move forged images; (b), (h) Color LBP images; (c), (i) The results based on DCT^[4]; (d), (j) The results based on Zernike^[14]; (e), (k) The results based on LBP^[17]; (f), (l) The results based on our method. Where (a), (g) solid and dashed rectangles are copied and pasted regions; (c)~(f), (i)~(l) marked regions are detected forged regions.)

下载: 全尺寸图片幻灯片

图 9 造成假象类复制-粘贴篡改检测结果示例 ((a), (g) 造成假象类复制-粘贴篡改图像; (b), (h) 彩色LBP图像; (c), (i) DCT^[4]算法检测结果; (d), (j) Zernike^[14]算法检测结果; (e), (k) LBP^[17]算法检测结果; (f), (l) 本文算法检测结果.其中, (a), (g) 实线和虚线框分别表示复制和粘贴区域; (c)~(f), (i)~(l) 标记区域为算法检测篡改区域)

Fig. 9 The exemplar results on spurious copy-move forgery detection ((a), (g) Spurious copy-move forged images; (b), (h) Color LBP images; (c), (i) The results based on DCT^[4]; (d), (j) The results based on Zernike^[14]; (e), (k) The results based on LBP^[17]; (f), (l) The results based on our method. Where (a), (g) solid and dashed rectangles are copied and pasted regions; (c)~(f) marked regions are detected forged regions; (i)~(l) green regions are detected forged regions.)

下载: 全尺寸图片幻灯片

图 10 添加80 dB高斯噪声实验结果

Fig. 10 The results of adding Gaussian noise with 80 dB

下载: 全尺寸图片幻灯片

图 11 添加高斯模糊 ([3, 1]) 实验结果

Fig. 11 The results of adding Gaussian blur ([3, 1])

下载: 全尺寸图片幻灯片

图 12 添加压缩因子90的JPEG重压缩实验结果

Fig. 12 The results of adding JPEG compression with quality factor 90

下载: 全尺寸图片幻灯片

图 13 后处理操作ROC曲线图 ((a)~(c) 分别为添加不同程度的高斯白噪声、高斯模糊和JPEG重压缩的ROC曲线)

Fig. 13 The ROC curves of post-processing operations (Where (a)~(c) are adding different degrees of Gaussian noise, Gaussian blur, and JPEG compression.)

下载: 全尺寸图片幻灯片

图 14 存在相似目标的真实图像示例

Fig. 14 The exemplar of authentic images with similar but genuine objects

下载: 全尺寸图片幻灯片

表 1 彩色空间选择分析

Table 1 The analysis on color space choice

彩色空间	彩色LBP图像提取时间 (s)	TPR (%)	FPR (%)
RGB	4	95	9
LAB	5	94	11
HSV	5.5	95	10

下载: 导出CSV

表 2 灰度级别g_th选择

Table 2 The choice of gray level g_th

灰度级别$g_{th}$	特征维度	TPR (%)	FPR (%)
4	16	85	15
8	64	91	13
16	256	95	7
32	024	95	6

下载: 导出CSV

参考文献(22)

[1]	Fridrich A J, Soukal B D, Lukáš A J. Detection of copy-move forgery in digital images. In: Proceedings of the 2003 Digital Forensic Research Workshop. Cleveland, USA, 2003.
[2]	Wang X, Zhang X, Li Z, Wang S. A DWT-DCT based passive forensics method for copy-move attacks. In: Proceedings of the 3rd International Conference on Multimedia Information Networking and Security. Shanghai, China: IEEE, 2011. 304-308
[3]	Hu J, Zhang H, Gao Q, Huang H. An improved lexicographical sort algorithm of copy-move forgery detection. In: Proceedings of the 2nd International Conference on Networking and Distributed Computing. Beijing, China: IEEE, 2011. 23-27
[4]	Cao Y J, Gao T G, Li F, Yang Q T. A robust detection algorithm for copy-move forgery in digital images. Forensic Science International, 2012, 214(1-3): 33-43 doi: 10.1016/j.forsciint.2011.07.015
[5]	Popescu A C, Farid H. Exposing Digital Forgeries by Detecting Duplicated Image Regions, Technical Report 2004-515, Dartmouth College, USA, 2004.
[6]	骆伟祺, 黄继武, 丘国平.鲁棒的区域复制图像篡改检测技术.计算机学报, 2007, 30(11): 1998-2007 http://www.cnki.com.cn/Article/CJFDTOTAL-JSJX200711013.htm Luo Wei-Qi, Huang Ji-Wu, Qiu Guo-Ping. Robust detection of region-duplication forgery in digital image. Chinese Journal of Computers, 2007, 30(11): 1998-2007 http://www.cnki.com.cn/Article/CJFDTOTAL-JSJX200711013.htm
[7]	Muhammad G, Hussain M, Bebis G. Passive copy move image forgery detection using undecimated dyadic wavelet transform. Digital Investigation, 2012, 9(1): 49-57 doi: 10.1016/j.diin.2012.04.004
[8]	Bayram S, Sencar H T, Memon N. An efficient and robust method for detecting copy-move forgery. In: Proceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal Processing. Taipei, China: IEEE, 2009. 1053-1056
[9]	Zhong J L, Gan Y F. Detection of copy-move forgery using discrete analytical Fourier-Mellin transform. Nonlinear Dynamics, 2016, 84(1): 189-202 doi: 10.1007/s11071-015-2374-9
[10]	Zhang T, Wang R. Copy-move forgery detection based on SVD in digital image. In: Proceedings of the 2nd International Congress on Image and Signal Processing. Tianjin, China: IEEE, 2009. 1-5
[11]	Liu G J, Wang J W, Lian S G, Wang Z Q. A passive image authentication scheme for detecting region-duplication forgery with rotation. Journal of Network and Computer Applications, 2011, 34(5): 1557-1565 doi: 10.1016/j.jnca.2010.09.001
[12]	Mahdian B, Saic S. Detection of copy-move forgery using a method based on blur moment invariants. Forensic Science International, 2007, 171(2-3): 180-189 doi: 10.1016/j.forsciint.2006.11.002
[13]	Ryu S J, Lee M J, Lee H K. Detection of copy-rotate-move forgery using Zernike moments. In: Proceedings of the 12th International Conference on Information Hiding. Calgary, AB, Canada: Springer, 2010. 51-65
[14]	Ryu S J, Kirchner M, Lee M J, Lee H K. Rotation invariant localization of duplicated image regions based on Zernike moments. IEEE Transactions on Information Forensics and Security, 2013, 8(8): 1355-1370 doi: 10.1109/TIFS.2013.2272377
[15]	Li L D, Li S S, Zhu H C, Wu X Y. Detecting copy-move forgery under affine transforms for image forensics. Computers and Electrical Engineering, 2014, 40(6): 1951-1962 doi: 10.1016/j.compeleceng.2013.11.034
[16]	Lynch G, Shih F Y, Liao H Y M. An efficient expanding block algorithm for image copy-move forgery detection. Information Sciences, 2013, 239: 253-265 doi: 10.1016/j.ins.2013.03.028
[17]	Davarzani R, Yaghmaie K, Mozaffari S, Tapak M. Copy-move forgery detection using multiresolution local binary patterns. Forensic Science International, 2013, 231(1-3): 61-72 doi: 10.1016/j.forsciint.2013.04.023
[18]	Cozzolino D, Poggi G, Verdoliva L. Efficient dense-field copy-move forgery detection. IEEE Transactions on Information Forensics and Security, 2015, 10(11): 2284-2297 doi: 10.1109/TIFS.2015.2455334
[19]	Porebski A, Vandenbroucke N, Macaire L. Haralick feature extraction from LBP images for color texture classification. In: Proceedings of the 1st Workshops on Image Processing Theory, Tools and Applications. Sousse: IEEE, 2008. 1-8
[20]	Haralick R M. Statistical and structural approaches to texture. Proceedings of the IEEE, 1979, 67(5): 786-804 doi: 10.1109/PROC.1979.11328
[21]	Amerini I, Ballan L, Caldelli R, Del Bimbo A, Serra G. A SIFT-based forensic method for copy-move attack detection and transformation recovery. IEEE Transactions on Information Forensics and Security, 2011, 6(3): 1099-1110 doi: 10.1109/TIFS.2011.2129512
[22]	Kakar P, Sudha N. Exposing post processed copy-paste forgeries through transform-invariant features. IEEE Transactions on Information Forensics and Security, 2012, 7(3): 1018-1028 doi: 10.1109/TIFS.2012.2188390