印鉴识别算法的研究
摘要
随着计算机技术的不断发展,伪造印章印鉴的难度大大下降,从而导致伪造印鉴引起的犯罪活动屡禁不止。印鉴识别算法的研究,其目标是以高技术水平进行科学的、准确的文件鉴定,向社会提供一个具有自动化的高技术保障。
本论文主要针对印鉴识别算法进行了研究。论文中首先分析了印鉴识别的难点和重点,然后针对这些难点和重点,提出了一种新的印鉴配准方法并提取了四种互补的特征。这种新的印鉴配准算法首先求出印鉴的几何中心,以几何中心为圆心,先在极径方向上抽样离散化,然后离散化每一个同心圆,重采样每个同心圆上的点,最后以小区域匹配求出旋转角度,完成印鉴的配准。实验表明,这种方法是有效的。
针对印鉴的特点,我们提取了四种不同的,互补的印鉴特征和各自的分类方法。这四种特征分别是:基于区域的形状特征提取、基于印鉴配准的特征提取、基于小波变换的纹理特征提取和改进的SIFT特征提取。基于区域的形状特征提取是利用Hu矩不变量提取印鉴的形状,然后对不同形状的印鉴进行分类;基于印鉴配准的特征提取是在印鉴配准的基础上,定义一个差异性函数,根据这个函数选取一个阈值判断印鉴的真伪;基于小波变换的纹理特征提取,主要是针对印鉴中的细纹理特征,依据低频子带的能量判断小波变换的级数,对小波变换后的子带图像提取均值和方差,作为特征向量输入支持向量机进行印鉴真伪的鉴别; SIFT是一种稳定的,匹配能力强的算法,针对SIFT实时性差,本论文提出了基于改进的SIFT特征提取,改进的SIFT是在印鉴配准的基础上,将标准印鉴和待测印鉴按角度等划分,使特征向量的搜索范围变小,从而减少匹配的时间。另外,根据印鉴识别的特点,印鉴识别取决于边缘点,背景点和非边缘点对印鉴的影响不大,我们将印鉴配准后再提取边缘,然后利用SIFT算法计算特征点,利用欧式距离分类匹配点。实验表明,改进的SIFT算法在匹配率和实时性上都要优于SIFT。
最后,本文介绍了多分类器融合的基础知识,采用投票规则将这四种不同的印鉴识别方法进行了融合。实验结果证明这四种特征是有效的,互补的。
With the development of the computer techniques, difficulties of forging seal decline. Therefore economic crimes caused by forgery seals are hard to be forbidden. The purpose of seal identification study is to provide a system that can identify tested seal images to be genuine or forgery, and it also can offer an automatic, efficient and accurate documentation authority identification way to avoid crimes.
This paper focuses on Seal identification. First, discuss the emphasis and difficulties of seal identification, Based on these emphasis and difficulties, a new registration method and four complementary features are used. This new registration method fit all kinds of seal. It first get the center of mass, then use the center of mass as center, make the radius and concentric circle discretization and resample the points in concentric circle, At last get the angle's value using small region matching and finish the registration. This method is proved efficient by experiment.
According to the feature of seal, this paper brings up four different and complementary features and their different classification methods. These features are shape based on Hu Moment Invariants, difference based on seal registration, texture based on wavelet and ISIFT. Shape features are used to classify the seal of different shape by abstracting Hu Moment. We define a difference function based on seal registration. According the difference function, we choose a threshold value to realize seal identification. Texture features based on wavelet choose mean and variance of all sub-channels as an input into SVM. SIFT is a stable and powerfully matching method. Because SIFT costs so much time, this paper gives an improved SIFT method to decrease the time cost. ISIFT divides the seal into several regions after registration and edge extracting. The small region makes searching cost much less time. In addition, the points on edge are key factor to determine seal identification, we can ignore other points. Then use Euclidean distance to judge feature points after registration and edge extracting. The experiment tells ISIFT is better than SIFT in aspect of rate of registration and real-time.
本论文主要针对印鉴识别算法进行了研究。论文中首先分析了印鉴识别的难点和重点,然后针对这些难点和重点,提出了一种新的印鉴配准方法并提取了四种互补的特征。这种新的印鉴配准算法首先求出印鉴的几何中心,以几何中心为圆心,先在极径方向上抽样离散化,然后离散化每一个同心圆,重采样每个同心圆上的点,最后以小区域匹配求出旋转角度,完成印鉴的配准。实验表明,这种方法是有效的。
针对印鉴的特点,我们提取了四种不同的,互补的印鉴特征和各自的分类方法。这四种特征分别是:基于区域的形状特征提取、基于印鉴配准的特征提取、基于小波变换的纹理特征提取和改进的SIFT特征提取。基于区域的形状特征提取是利用Hu矩不变量提取印鉴的形状,然后对不同形状的印鉴进行分类;基于印鉴配准的特征提取是在印鉴配准的基础上,定义一个差异性函数,根据这个函数选取一个阈值判断印鉴的真伪;基于小波变换的纹理特征提取,主要是针对印鉴中的细纹理特征,依据低频子带的能量判断小波变换的级数,对小波变换后的子带图像提取均值和方差,作为特征向量输入支持向量机进行印鉴真伪的鉴别; SIFT是一种稳定的,匹配能力强的算法,针对SIFT实时性差,本论文提出了基于改进的SIFT特征提取,改进的SIFT是在印鉴配准的基础上,将标准印鉴和待测印鉴按角度等划分,使特征向量的搜索范围变小,从而减少匹配的时间。另外,根据印鉴识别的特点,印鉴识别取决于边缘点,背景点和非边缘点对印鉴的影响不大,我们将印鉴配准后再提取边缘,然后利用SIFT算法计算特征点,利用欧式距离分类匹配点。实验表明,改进的SIFT算法在匹配率和实时性上都要优于SIFT。
最后,本文介绍了多分类器融合的基础知识,采用投票规则将这四种不同的印鉴识别方法进行了融合。实验结果证明这四种特征是有效的,互补的。
With the development of the computer techniques, difficulties of forging seal decline. Therefore economic crimes caused by forgery seals are hard to be forbidden. The purpose of seal identification study is to provide a system that can identify tested seal images to be genuine or forgery, and it also can offer an automatic, efficient and accurate documentation authority identification way to avoid crimes.
This paper focuses on Seal identification. First, discuss the emphasis and difficulties of seal identification, Based on these emphasis and difficulties, a new registration method and four complementary features are used. This new registration method fit all kinds of seal. It first get the center of mass, then use the center of mass as center, make the radius and concentric circle discretization and resample the points in concentric circle, At last get the angle's value using small region matching and finish the registration. This method is proved efficient by experiment.
According to the feature of seal, this paper brings up four different and complementary features and their different classification methods. These features are shape based on Hu Moment Invariants, difference based on seal registration, texture based on wavelet and ISIFT. Shape features are used to classify the seal of different shape by abstracting Hu Moment. We define a difference function based on seal registration. According the difference function, we choose a threshold value to realize seal identification. Texture features based on wavelet choose mean and variance of all sub-channels as an input into SVM. SIFT is a stable and powerfully matching method. Because SIFT costs so much time, this paper gives an improved SIFT method to decrease the time cost. ISIFT divides the seal into several regions after registration and edge extracting. The small region makes searching cost much less time. In addition, the points on edge are key factor to determine seal identification, we can ignore other points. Then use Euclidean distance to judge feature points after registration and edge extracting. The experiment tells ISIFT is better than SIFT in aspect of rate of registration and real-time.
引文
[1]顾惠敏.印鉴欣赏与收藏[M].上海:上海书店出版社,1996
[2]蔡国声.印章三千年[M].上海:上海文化出版社,1996
[3]刘一闻.印章[M].上海:上海古籍出版社,1995
[4]孙尉祖.印章[M].上海:上海人民美术出版社,1998
[5]中华人民共和国票据法.http://www.china.com.cn/chinese/law/647634.htm,2004 -8-9/2007-4-6
[6]K.Ueda ,Y.Nakamura.Automatic Verification of Seal Impression Pattern[J].Proc.7th.Int. Conf.On Recognition,1984,pp.1019-1021
[7]Fan T. H. , Tsai W.H..Automatic Chinese Seal Identification[J].Computer Vision Graphics and Image Processing, 1984, 25:311-330
[8]Takahiko Horiuchi.,Automatic Seal Verification by Evaluating Positive Cost[J]. ICDAR 2001: 572-576
[9]胡庆,杨静宇等.基于知识的印鉴鉴别方法[J].自动化学报,1991, 17(6):696-703
[10]S Lee,J H Kim.Unconstrained seal imprint verification using attributed stroke graph matching[J].Pattern Recognition, 1989,22(6):653-664
[11]曾东武,刘重庆等.Fourier描述符的印鉴识别方法[J].上海交通大学学报,1998,32 (9) : 43-45
[12]刘丽珍,傅德胜.基于Bubble小波函数的印鉴鉴别[J].南京气象学院学报,2002,25(5):698-701
[13]张先萌,罗安玉等.用矩不变量实现印鉴自动识别[J].电子学报,1995,23(4):100-103
[14]孟章荣.各种颜色模型选用需求分析[J].中国图像图形学报.1996.1(3):238一241
[15]李弼程,彭天强等.智能图像处理技术[M].北京:电子工业出版社,2004
[16] Mallat, S.G.A theory for multiresolution signal decomposition: the wavelet representation[J]. IEEE Trans Pattern Analysis and Machine Intelligence. 1989.11(7): 674-693.
[17] Kenneth R.Castleman著,朱志刚等译.数字图像处理[M].北京:电子工业出版社,2002.2
[18]高文,董胜富,周世意.基于边缘匹配的印鉴自动鉴别方法.模式识别与人工智能,1994, 7(4): 338-342
[19]张登银,汤志云.基于边缘配准的印鉴自动鉴别方法.南京邮电学院学报:白然科学版,2003,23(3):25-29
[20].L G Robert.Machine perception of three-dimensional solids[J].In:Optical and Electro-Optical Information Processing,Tippett J,et al.eds.,1965,pp:159-197
[21]Canny J.A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,PAMI-8(6):679-698
[22]D.Marr,E C Hildreth.Theory of edge detection[J].Proc.Royal Soc. Of London,1980,B207: 187-217
[23]谷口庆治编,朱虹、廖学成、乐静等译.数字图像处理-基础篇[M].北京:科学出版社,2002
[24]Hu M K.Visual pattern Recognition by Moment Invariants[J].IRE Transactions on Information Theory,1962,IT-8:179-187.
[25]Arvazhagan S, Ganesan L. Texture Classification Using Wavelet Transform[J]. Pattern Recognition Letters, 2003, 24(10): 1513-1521
[26]Mallat S. Characterization of Signals from Multiscale Edges[J]. IEEE Trans. on Pattern Anal. Machine Intell., 1992, 14(7): 710-732
[27]Mallat S. A Theory for Multiresolution Signal Decomposition: The Wavelet Representation[J]. IEEE Trans. on Pattern Anal. Machine Intell., 1989, 11(5): 674-693
[28]古红英,潘云鹤.基于纹理特征融合约减的虹膜识别研究[J].计算机研究与发展.2006,43(1):126-131
[29]王耀南,李树涛,毛建旭等.计算机图像处理与识别技术[M].北京:高等教育出版社,2001
[30]王孝国,张小宁等.一种基于离散小波变换和支持向量机的人脸识别新方法[J] .解放军理工大学学报(自然科学版),2006,7(6):515-519
[31]杨凌霄,王成硕等.基于小波变换和支持向量机的人脸识别方法研究[J].自动化技术与应用,2007,26(4):52-55
[32]汪妍,夏良正.基于小波变换的人脸特征定位[J].东南大学学报(自然科学版), 2004,S1: 118-121
[33]杨友生,王耀明,郭世杰.基于小波变换和支持向量机的人脸检测[J].微机发展, 2004,11: 66-68
[34]雷寒生,朱小平.基于小波变换和矢量量化的人脸图像压缩[J].中国图象图形学报,2002,7(A)(1):44-49.
[35]耿建玲,陈纯,陈耀武等.基于小波分析和SVM的印鉴图像识别[J].计算机工程与应用,2006,19:194-197.
[36]O Chapelle,V Vapnik et al.Choosing multiple parameters for support vector machines[J].Machine Learning,2002,46:131-159
[37]David G. Lowe. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004 60(2):91-110
[38]Mikolajczyk K,Schmid C.A performance evaluation of local descriptors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(10):1615-1630
[39]Yan Ke, Rahul Sukthankar. PCA-SIFT: A More Distinctive Representation for Local Image Descriptors[J]. CVPR (2) 2004: 506-513
[40]Besl P,Jain R.Segmentation and classification of range images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1987,9:608-620
[41]I. Stamos, P. E. Allen, 3-D model construction using range and image data[J]. CVPR00(I: 531-536).
[42]Besl P J,McKay N D.A Method for Registration of 3-D Shapes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(2):239-256
[43]Y. Yu, A. Ferencz, J. Malik. Extracting Objects from Range and Radiance Images[J]. IEEE Trans, Visualization and Computer Graphics, 2001,7
[44]Faugeras O,Robert L.What can two images tell us about the third one[C]. Proceedings of the Europe Conference on Computer Vision, Sweden, 1994
[45]Koenderink J.The structure of images[J]. Biological Cybernetics, 1984, 50: 363-396.
[46] Lindeberg T.Scale-Space for discrete Signals[J]. IEEE Transactions PAM I, 1980, 207: 187-217
[47]Babaud J , Witkin A P, Baudin Metal. Uniqueness of the Gaussian kernel for scale - space filtering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996, 8(1)26-33
[48]A.A.SPanjersberg. Combination of different systems of the Recognition of handwritten digits. In:2nd International Joint Conference on Pattern Recognition,copenhagen,PP.208-209,1974
[49]L.Lam and C.Y.Suen.Optimal combinations of Pattern classifiers.Pattern Recognition Letters,vol.19,no.9,PP.945-954,1995
[50]C.Dietrich,F.Schwenker,K.Riede,and G.Palm.Classification of Bioacoustic time series utilizing Pulse detection,time and frequency Features and data fusion.2001
[51]Y.S.Huang and C.Y.Suen.A method of combining multiple Experts of the recognition of unconstrained handwritten numerals.IEEE Transactions On Pattern Analysis and Machine Intelligence,vol.17,no.1,PP.90-94,1995
[2]蔡国声.印章三千年[M].上海:上海文化出版社,1996
[3]刘一闻.印章[M].上海:上海古籍出版社,1995
[4]孙尉祖.印章[M].上海:上海人民美术出版社,1998
[5]中华人民共和国票据法.http://www.china.com.cn/chinese/law/647634.htm,2004 -8-9/2007-4-6
[6]K.Ueda ,Y.Nakamura.Automatic Verification of Seal Impression Pattern[J].Proc.7th.Int. Conf.On Recognition,1984,pp.1019-1021
[7]Fan T. H. , Tsai W.H..Automatic Chinese Seal Identification[J].Computer Vision Graphics and Image Processing, 1984, 25:311-330
[8]Takahiko Horiuchi.,Automatic Seal Verification by Evaluating Positive Cost[J]. ICDAR 2001: 572-576
[9]胡庆,杨静宇等.基于知识的印鉴鉴别方法[J].自动化学报,1991, 17(6):696-703
[10]S Lee,J H Kim.Unconstrained seal imprint verification using attributed stroke graph matching[J].Pattern Recognition, 1989,22(6):653-664
[11]曾东武,刘重庆等.Fourier描述符的印鉴识别方法[J].上海交通大学学报,1998,32 (9) : 43-45
[12]刘丽珍,傅德胜.基于Bubble小波函数的印鉴鉴别[J].南京气象学院学报,2002,25(5):698-701
[13]张先萌,罗安玉等.用矩不变量实现印鉴自动识别[J].电子学报,1995,23(4):100-103
[14]孟章荣.各种颜色模型选用需求分析[J].中国图像图形学报.1996.1(3):238一241
[15]李弼程,彭天强等.智能图像处理技术[M].北京:电子工业出版社,2004
[16] Mallat, S.G.A theory for multiresolution signal decomposition: the wavelet representation[J]. IEEE Trans Pattern Analysis and Machine Intelligence. 1989.11(7): 674-693.
[17] Kenneth R.Castleman著,朱志刚等译.数字图像处理[M].北京:电子工业出版社,2002.2
[18]高文,董胜富,周世意.基于边缘匹配的印鉴自动鉴别方法.模式识别与人工智能,1994, 7(4): 338-342
[19]张登银,汤志云.基于边缘配准的印鉴自动鉴别方法.南京邮电学院学报:白然科学版,2003,23(3):25-29
[20].L G Robert.Machine perception of three-dimensional solids[J].In:Optical and Electro-Optical Information Processing,Tippett J,et al.eds.,1965,pp:159-197
[21]Canny J.A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,PAMI-8(6):679-698
[22]D.Marr,E C Hildreth.Theory of edge detection[J].Proc.Royal Soc. Of London,1980,B207: 187-217
[23]谷口庆治编,朱虹、廖学成、乐静等译.数字图像处理-基础篇[M].北京:科学出版社,2002
[24]Hu M K.Visual pattern Recognition by Moment Invariants[J].IRE Transactions on Information Theory,1962,IT-8:179-187.
[25]Arvazhagan S, Ganesan L. Texture Classification Using Wavelet Transform[J]. Pattern Recognition Letters, 2003, 24(10): 1513-1521
[26]Mallat S. Characterization of Signals from Multiscale Edges[J]. IEEE Trans. on Pattern Anal. Machine Intell., 1992, 14(7): 710-732
[27]Mallat S. A Theory for Multiresolution Signal Decomposition: The Wavelet Representation[J]. IEEE Trans. on Pattern Anal. Machine Intell., 1989, 11(5): 674-693
[28]古红英,潘云鹤.基于纹理特征融合约减的虹膜识别研究[J].计算机研究与发展.2006,43(1):126-131
[29]王耀南,李树涛,毛建旭等.计算机图像处理与识别技术[M].北京:高等教育出版社,2001
[30]王孝国,张小宁等.一种基于离散小波变换和支持向量机的人脸识别新方法[J] .解放军理工大学学报(自然科学版),2006,7(6):515-519
[31]杨凌霄,王成硕等.基于小波变换和支持向量机的人脸识别方法研究[J].自动化技术与应用,2007,26(4):52-55
[32]汪妍,夏良正.基于小波变换的人脸特征定位[J].东南大学学报(自然科学版), 2004,S1: 118-121
[33]杨友生,王耀明,郭世杰.基于小波变换和支持向量机的人脸检测[J].微机发展, 2004,11: 66-68
[34]雷寒生,朱小平.基于小波变换和矢量量化的人脸图像压缩[J].中国图象图形学报,2002,7(A)(1):44-49.
[35]耿建玲,陈纯,陈耀武等.基于小波分析和SVM的印鉴图像识别[J].计算机工程与应用,2006,19:194-197.
[36]O Chapelle,V Vapnik et al.Choosing multiple parameters for support vector machines[J].Machine Learning,2002,46:131-159
[37]David G. Lowe. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004 60(2):91-110
[38]Mikolajczyk K,Schmid C.A performance evaluation of local descriptors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(10):1615-1630
[39]Yan Ke, Rahul Sukthankar. PCA-SIFT: A More Distinctive Representation for Local Image Descriptors[J]. CVPR (2) 2004: 506-513
[40]Besl P,Jain R.Segmentation and classification of range images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1987,9:608-620
[41]I. Stamos, P. E. Allen, 3-D model construction using range and image data[J]. CVPR00(I: 531-536).
[42]Besl P J,McKay N D.A Method for Registration of 3-D Shapes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(2):239-256
[43]Y. Yu, A. Ferencz, J. Malik. Extracting Objects from Range and Radiance Images[J]. IEEE Trans, Visualization and Computer Graphics, 2001,7
[44]Faugeras O,Robert L.What can two images tell us about the third one[C]. Proceedings of the Europe Conference on Computer Vision, Sweden, 1994
[45]Koenderink J.The structure of images[J]. Biological Cybernetics, 1984, 50: 363-396.
[46] Lindeberg T.Scale-Space for discrete Signals[J]. IEEE Transactions PAM I, 1980, 207: 187-217
[47]Babaud J , Witkin A P, Baudin Metal. Uniqueness of the Gaussian kernel for scale - space filtering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996, 8(1)26-33
[48]A.A.SPanjersberg. Combination of different systems of the Recognition of handwritten digits. In:2nd International Joint Conference on Pattern Recognition,copenhagen,PP.208-209,1974
[49]L.Lam and C.Y.Suen.Optimal combinations of Pattern classifiers.Pattern Recognition Letters,vol.19,no.9,PP.945-954,1995
[50]C.Dietrich,F.Schwenker,K.Riede,and G.Palm.Classification of Bioacoustic time series utilizing Pulse detection,time and frequency Features and data fusion.2001
[51]Y.S.Huang and C.Y.Suen.A method of combining multiple Experts of the recognition of unconstrained handwritten numerals.IEEE Transactions On Pattern Analysis and Machine Intelligence,vol.17,no.1,PP.90-94,1995