基于多方向互交叉模式算子与极坐标变换的鲁棒图像哈希算法
|
摘要
为了提高哈希序列对几何攻击的鲁棒性与正确识别率,设计了基于多方向互交叉模式算子与极坐标变换的鲁棒哈希算法。引入插值运算与Gaussian滤波器,完成图像的尺寸规范化与去噪处理,使其对于任意的可疑目标均可输出一个固定长度的哈希序列。基于极坐标变换(log-polar transform,LPT),对滤波规范图像实施处理,输出抗旋转攻击的二次图像。随后,利用多方向互交叉模式算子,从8个方向将二次图像变换为两个编码映射。将两个编码映射分割为非重叠子块,通过提取这些子块的直方图,将其视为纹理特征,作为第一个哈希序列。利用强度概率密度梯度代替强度梯度,对SURF (speeded up robust features)方法予以改进,充分提取图像中的稳定角点,形成角点图像;将角点图像分割为一系列的非重叠子块,通过计算每个子块所含的角点数量,将含有结构信息最丰富的子块予以标记,输出其在图像中对应的位置信息;并借助离散小波变换(discrete wave transform,DWT)来分解这些标记子块,获取每个子块对应的低频系数。联合位置信息与低频系数,形成结构特征,作为第二个哈希序列。设计加密机制,分别对两个哈希序列完成扩散,从而形成最终的哈希序列。通过计算源图像与可疑图像之间的l_2范数距离,将其与用户识别阈值的大小对比,对目标的真实性完成判别。试验数据显示:较已有的哈希生成机制而言,所提方法拥有更高的鲁棒性,对各类几何攻击均有更高的识别准确率。
In order to improve the robustness of hash sequences against geometric attacks and correct recognition rate,a robust hash algorithm based on multi-directional mutual cross pattern operator and polar coordinate transformation is designed. By introducing interpolation operation and gaussian filter,the size normalization and denoising of the image are completed to output a fixed-length hash sequence for any suspicious target. And the filtered standard image is processed by using the polar coordinate transformation to output the secondary image which can against rotation attack. Subsequently,the secondary image is transformed into two coding maps from eight directions by using the multi-directional mutual cross pattern operator. Two coding maps are segmented into non-overlapping subblocks,and the histograms of these sub-blocks are extracted to treat as texture features as the first hash sequence.The SURF method is improved by replacing the intensity gradient with the intensity probability density gradient to fully extract the stable corners in the image for forming the corner image. The corner image is divided into a series of non-overlapping sub-blocks,and the sub-blocks with the richest structural information are marked by calculating the number of corners contained in each sub-block to output their corresponding position information in the image.And the low frequency coefficients corresponding to each sub-block are obtained by decomposing these marker subblocks with the help of discrete wavelet transform. The structural features are formed by combining location information with low frequency coefficients,which as the second hash sequence. The encryption mechanism is designed to diffuse the two hash sequences for forming the final hash sequence. By calculating the l_2 norm distance between the source image and the suspicious image,the authenticity of the target is judged based on comparing it with the user recognition threshold. The experimental data show that the proposed method has higher Robustness and recognition accuracy for various geometric attacks than the existing hash generation mechanism.
In order to improve the robustness of hash sequences against geometric attacks and correct recognition rate,a robust hash algorithm based on multi-directional mutual cross pattern operator and polar coordinate transformation is designed. By introducing interpolation operation and gaussian filter,the size normalization and denoising of the image are completed to output a fixed-length hash sequence for any suspicious target. And the filtered standard image is processed by using the polar coordinate transformation to output the secondary image which can against rotation attack. Subsequently,the secondary image is transformed into two coding maps from eight directions by using the multi-directional mutual cross pattern operator. Two coding maps are segmented into non-overlapping subblocks,and the histograms of these sub-blocks are extracted to treat as texture features as the first hash sequence.The SURF method is improved by replacing the intensity gradient with the intensity probability density gradient to fully extract the stable corners in the image for forming the corner image. The corner image is divided into a series of non-overlapping sub-blocks,and the sub-blocks with the richest structural information are marked by calculating the number of corners contained in each sub-block to output their corresponding position information in the image.And the low frequency coefficients corresponding to each sub-block are obtained by decomposing these marker subblocks with the help of discrete wavelet transform. The structural features are formed by combining location information with low frequency coefficients,which as the second hash sequence. The encryption mechanism is designed to diffuse the two hash sequences for forming the final hash sequence. By calculating the l_2 norm distance between the source image and the suspicious image,the authenticity of the target is judged based on comparing it with the user recognition threshold. The experimental data show that the proposed method has higher Robustness and recognition accuracy for various geometric attacks than the existing hash generation mechanism.
引文
1 Mathon B,Cayre F,Bas P.Optimal transport for secure spread-spectrum watermarking of still images[J].IEEE Trans Image Process,2014,23(4):1694-1705
2王彦超.基于压缩量化与邻域空间LBP算子的图像哈希算法[J].包装工程,2017,38(21):191-198Wang Y C.Image hashing algorithms based on compressed quantization and neighborhood space LBP operator[J].Packaging Engineering,2017,38(21):191-198
3 Qu W,Wang D L,Feng S.A novel cross-modal hashing algorithm based on multimodal deep learning[J].Science China Information Sciences,2017,60(9):156-167
4王彦超,郭静博,周丽宴.基于数据投影降维机制与对称局部二值模式的紧凑图像哈希算法[J].激光与光电子学进展,2017,54(2):021004-021016Wang Y C,Guo J B,Zhou L Y.Compact image hashing algorithm based on data projection dimensionality reduction mechanism and symmetrical local binary mode[J].Progress in Laser and Optoelectronics,2017,54(2):021004-021016
5王彦超.基于邻域LBP算子与块截断编码的图像哈希算法[J].计算机工程与设计,2018,39(7):2027-2035Wang Y C.Image hashing algorithm based on neighborhood LBP operator and block truncation coding[J].Computer Engineering and Design,2018,39(7):2027-2035
6 Pun C M,Yan C P,Yuan X C.Robust image hashing using progressive feature selection for tampering detection[J].Multimedia Tools&Applications,2018,77(10):11609-11633
7 Qin C,Sun M H,Chang C C.Perceptual hashing for color images based on hybrid extraction of structural features[J].Signal Processing,2018,142(8):194-205
8 Adjed F,Gardezi S,Ababsa F.Fusion of structural and textural features for melanoma recognition[J].IET Computer Vision,2018,12(7):185-195
9金晓民,张丽萍.混合特征与颜色矢量角度的图像哈希认证算法[J].计算机科学与探索,2018,38(7):1102-1115Jin X M,Zhang L P.Image hash authentication algorithms based on mixed features and color vector angle[J].Computer Science and Exploration,2018,38(7):1102-1115
10 Sandoz M F,Nour A S.Robust copy-move forgery revealing in digital images using polar coordinate system[J].Neurocomputing,2017,265(11):57-65
11 Ding C X,Choi J Y,Tao D C.Multi-directional multi-level dual-cross patterns for robust face recognition[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2016,38(3):518-531
12杨金劳,付利军,张福泉.基于椭圆特征区域与重要位平面分解的鲁棒图像水印算法[J].包装工程,2018,39(21):206-215Yang J L,Fu L J,Zhang F Q.Robust image watermarking based on elliptic feature region and important bit plane decomposition[J].Packaging Engineering,2018,39(21):206-215
13 Wang X Y,Liu Y N,Li S.A new robust digital watermarking using local polar harmonic transform[J].Computer and Electrical Engineering,2015,46(5):403-418
14孙达.基于概率密度的图像处理算法的研究与应用[D].哈尔滨:哈尔滨工业大学,2009:28-35Sun D.Research and application of image processing algorithms based on probability density[D].Harbin:Harbin University of Technology,2009:28-35
15 Wang X Y,Niu P P,Yang H Y.Affine invariant image watermarking using intensity probability density-based Harris laplace detector[J].Journal of Visual Communication and Image Representation,2012,23(6):892-907
16 Makbol N M,Khoo B E,Rassem,T H.Block-based discrete wavelet transform-singular value decomposition image watermarking scheme using human visual system characteristics[J].IET Image Processing,2016,10(1):34-52
17 Li C Q,Xie T,Liu Q.Cryptanalyzing image encryption using chaotic logistic map[J].Nonlinear Dynamics,2014,78(2):1545-1551
18 Ye Q L,Fu L Y,Zhang Z.Lp-and ls-norm distance based robust linear discriminant analysis[J].Neural Networks,2018,39(8):1017-2028
19 Schaefer G,Stich M.UCID-an uncompressed colour image database[C]//Proceedings of SPIE,Storage and Retrieval Methods and Applications for Multi-media.Bellingham:SPIE,2004:472-480
2王彦超.基于压缩量化与邻域空间LBP算子的图像哈希算法[J].包装工程,2017,38(21):191-198Wang Y C.Image hashing algorithms based on compressed quantization and neighborhood space LBP operator[J].Packaging Engineering,2017,38(21):191-198
3 Qu W,Wang D L,Feng S.A novel cross-modal hashing algorithm based on multimodal deep learning[J].Science China Information Sciences,2017,60(9):156-167
4王彦超,郭静博,周丽宴.基于数据投影降维机制与对称局部二值模式的紧凑图像哈希算法[J].激光与光电子学进展,2017,54(2):021004-021016Wang Y C,Guo J B,Zhou L Y.Compact image hashing algorithm based on data projection dimensionality reduction mechanism and symmetrical local binary mode[J].Progress in Laser and Optoelectronics,2017,54(2):021004-021016
5王彦超.基于邻域LBP算子与块截断编码的图像哈希算法[J].计算机工程与设计,2018,39(7):2027-2035Wang Y C.Image hashing algorithm based on neighborhood LBP operator and block truncation coding[J].Computer Engineering and Design,2018,39(7):2027-2035
6 Pun C M,Yan C P,Yuan X C.Robust image hashing using progressive feature selection for tampering detection[J].Multimedia Tools&Applications,2018,77(10):11609-11633
7 Qin C,Sun M H,Chang C C.Perceptual hashing for color images based on hybrid extraction of structural features[J].Signal Processing,2018,142(8):194-205
8 Adjed F,Gardezi S,Ababsa F.Fusion of structural and textural features for melanoma recognition[J].IET Computer Vision,2018,12(7):185-195
9金晓民,张丽萍.混合特征与颜色矢量角度的图像哈希认证算法[J].计算机科学与探索,2018,38(7):1102-1115Jin X M,Zhang L P.Image hash authentication algorithms based on mixed features and color vector angle[J].Computer Science and Exploration,2018,38(7):1102-1115
10 Sandoz M F,Nour A S.Robust copy-move forgery revealing in digital images using polar coordinate system[J].Neurocomputing,2017,265(11):57-65
11 Ding C X,Choi J Y,Tao D C.Multi-directional multi-level dual-cross patterns for robust face recognition[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2016,38(3):518-531
12杨金劳,付利军,张福泉.基于椭圆特征区域与重要位平面分解的鲁棒图像水印算法[J].包装工程,2018,39(21):206-215Yang J L,Fu L J,Zhang F Q.Robust image watermarking based on elliptic feature region and important bit plane decomposition[J].Packaging Engineering,2018,39(21):206-215
13 Wang X Y,Liu Y N,Li S.A new robust digital watermarking using local polar harmonic transform[J].Computer and Electrical Engineering,2015,46(5):403-418
14孙达.基于概率密度的图像处理算法的研究与应用[D].哈尔滨:哈尔滨工业大学,2009:28-35Sun D.Research and application of image processing algorithms based on probability density[D].Harbin:Harbin University of Technology,2009:28-35
15 Wang X Y,Niu P P,Yang H Y.Affine invariant image watermarking using intensity probability density-based Harris laplace detector[J].Journal of Visual Communication and Image Representation,2012,23(6):892-907
16 Makbol N M,Khoo B E,Rassem,T H.Block-based discrete wavelet transform-singular value decomposition image watermarking scheme using human visual system characteristics[J].IET Image Processing,2016,10(1):34-52
17 Li C Q,Xie T,Liu Q.Cryptanalyzing image encryption using chaotic logistic map[J].Nonlinear Dynamics,2014,78(2):1545-1551
18 Ye Q L,Fu L Y,Zhang Z.Lp-and ls-norm distance based robust linear discriminant analysis[J].Neural Networks,2018,39(8):1017-2028
19 Schaefer G,Stich M.UCID-an uncompressed colour image database[C]//Proceedings of SPIE,Storage and Retrieval Methods and Applications for Multi-media.Bellingham:SPIE,2004:472-480