摘要
针对数字图像中复制-移动伪造的检测,提出一种结合灰度差分统计法(GDS)特征提取和多层感知机(MLP)神经网络分类器的检测方法。首先,将图像转换为灰度图像,并获得图像的灰度差分矩阵。然后,根据灰度分布、距离分布等信息,利用灰度差分统计法构建5个特征构成特征向量。最后,基于提取的特征,通过MLP神经网络进行分类,以此来检测该图像块是否为伪造区域。在实验中,将提出的GDS特征与传统SIFT特征进行了比较,结果表明,该方法能够有效检测出伪造区域,具有较高的准确性。
For the detection of copy-move forgery in digital images,a detection method is proposed based on gray-level difference statistics(GDS)feature extraction and multi-layer perceptron(MLP)neural network classifier.Firstly,the image is converted to a gray image and the gray difference matrix of the image is obtained.Then,based on the gray-level difference distribution and distance distribution,five features are extracted by using the gray-level difference statistics to form an eigenvector.Finally,based on the extracted features,classification is performed by the MLP neural network to detect whether the image block is a forged region.In the experiment,the proposed GDS features are compared with the traditional SIFT features.The results show that this method can effectively detect fake regions with high accuracy.
引文
[1]夏伟,王萍,马佳艳,等.基于图像信息测度和极限学习机的图像边缘检测[J].微型电脑应用,2017,33(11):35-39.
[2]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.
[3]李晓飞,李鹏飞.基于SIFT的伪造图像盲检测算法[J].长春大学学报:自然科学版,2014,24(10):1354-1357.
[4]Tu K H,Huynh K V,Le-Tien T,et al.A survey on Image Forgery Detection techniques[C]//IEEE Rivf International Conference on Computing&Communication Technologies-Research,Innovation,and Vision for the Future.IEEE,2015:71-76.
[5]Li L,Li S,Zhu H,et al.An efficient scheme for detecting copy-move forged images by local binary patterns[J].Journal of Information Hiding&Multimedia Signal Processing,2013,4(1):46-56.
[6]Yang J,Ran P,Xiao D,et al.Digital image forgery forensics by using undecimated dyadic wavelet transform and Zernike moments[J].Journal of Computational Information Systems,2013,9(16):6399-6408.
[7]Kashyap A,Joshi S D.Detection of copy-move forgery using wavelet decomposition[C]//International Conference on Signal Processing and Communication.IEEE,2014:396-400.
[8]刘辉,张云生,张印辉,等.基于灰度差分统计的火焰图像纹理特征提取[J].控制工程,2013,20(2):213-218.
[9]万雪,张祖勋,柯涛.一种利用零交叉点理论的改进SIFT特征提取算法[J].武汉大学学报(信息科学版),2013,38(3):270-273.
[10]Rajkumar R,Singh K M.Digital image forgery detection using SIFT feature[C]//International Symposium on Advanced Computing and Communication.IEEE,2016:186-191.
[11]Hsu C Y,Lu C S,Pei S C.Image feature extraction in encrypted domain with privacy-preserving SIFT[J].IEEE Trans Image Process,2012,21(11):4593-4607.
[12]李传朋,秦品乐,张晋京.基于深度卷积神经网络的图像去噪研究[J].计算机工程,2017,43(3):253-260.
[13]Sambhaji E S.Leaf Recognition Algorithm Using MLP Neural Network Based Image Processing[J].International Journal of Engineering Sciences&Research Technology,2014,3(5):132-137.