双频感应测井仪接收机前端电路设计
|
摘要
印章作为中国社会信用体系的一种表现形式，在日常生活中发挥着非常重要的作用。随着科技的不断发展，印章制作工艺得到了很大的提高。从传统人工雕刻发展到了机械雕刻、激光雕刻等先进的制作方式。随即带来的是假冒伪劣印章在识别中的难度大大的提高，传统人工比对方法的有效性大大降低，因此出现了计算机印章自动鉴定研究。印章自动鉴定属于图像模式识别领域，其主要分为印文分割、印文配准、特征提取及特征识别几个步骤。到目前为止，已有一些很好的研究成果被应用到实际工作中。但由于印章图像的复杂性，尚有一些问题未有效解决。
本文通过研究，提出了一种新的方法，成功的将脉冲耦合神经网络(PCNN)应用到印章自动识别中。利用PCNN的自动波的特性，可有效的定位印文在图像中的位置。在获得了印文所在位置后，可获得印文图像上的关键有效像素点的色彩等信息，这样就能够得到最为准确的阈值，可有效的将印文从复杂背景中分割出来。此外还提出了将主元成分分析(PCA)神经网络方法应用到印文配准的工作中。PCA方法可以通过全局分布特点，快速的将印文图像的大致方向确定，并将印文基本配准，然后再基于基本配准的印文图像，进行小范围的精确搜索，达到最佳的精确配准效果。该方法同样得到了较好的结果。
为了验证PCNN及PCA方法在实际应用中的效果，本文通过将这些方法结合实际印章图像用MATLAB进行了仿真试验。通过试验表明，本文提出的结合PCNN的方法可准确的将有效印文从复杂背景中分割出来。同时，通过与PCA结合的方法可非常精确的将印文图像进行配准，对后续的特征提取及识别工作提供了很好的支持。进一步，为了将研究成果紧密的结合到真实的应用环境中，本文设计并实现了智能印章识别系统(IntelliSIS)，本文对IntelliSIS系统的设计框架、系统实现、系统性能等作了详细介绍。
As the form of credit system of Chinese society, seal is very important for this. By the development of technology, the producing techniques have a great improvement. With this improvement, the difficulty of fake seal recognition is also being higher and higher and the validity of traditional manual comparison is greatly low. Thus the automatic computer identification research arose. Automatic seal identification belongs to pattern recognition domain and includes seal segmentation, seal registration, feature extraction and feature identification these steps. Recently, some good research results have been applied in practice. But because of the complexity of seal images, there are still many problems need to be resolved.
According to research, we have proposed a new method, applying the Pulse Couple Neural Network (PCNN) in automatic seal identification. By using the auto wave character of PCNN, this method can locate the seal in the image. After we get the location of the seal imprint we can obtain the available critical color information of the seal imprint itself and segment seal texture from the image efficiently by using the accurate threshold. At the same time we also applied the Principal Components Analysis (PCA) neural network in seal registration successfully. The PCA can obtain the approximate direction of the seal imprint by using the global distribution of the image pixels of seal imprint. After this it can use the local search method for precisely registration and get a very good result.
To validate the PCNN and PCA methods in practice, we did the simulation by using the real seal images and PCNN and PCA in MATLAB. Through the experiment, the PCNN method is efficient, it can segment seal texture from the whole seal image precisely and PCA method can also register the seals precisely. This will helpful to the subsequent steps. Furthermore, to integrate the methods in a practical applied environment, we have designed and implemented an Intelligent Seal Identification System (IntelliSIS), it will introduce the details of system framework, implementation and system performance in this article.
本文通过研究，提出了一种新的方法，成功的将脉冲耦合神经网络(PCNN)应用到印章自动识别中。利用PCNN的自动波的特性，可有效的定位印文在图像中的位置。在获得了印文所在位置后，可获得印文图像上的关键有效像素点的色彩等信息，这样就能够得到最为准确的阈值，可有效的将印文从复杂背景中分割出来。此外还提出了将主元成分分析(PCA)神经网络方法应用到印文配准的工作中。PCA方法可以通过全局分布特点，快速的将印文图像的大致方向确定，并将印文基本配准，然后再基于基本配准的印文图像，进行小范围的精确搜索，达到最佳的精确配准效果。该方法同样得到了较好的结果。
为了验证PCNN及PCA方法在实际应用中的效果，本文通过将这些方法结合实际印章图像用MATLAB进行了仿真试验。通过试验表明，本文提出的结合PCNN的方法可准确的将有效印文从复杂背景中分割出来。同时，通过与PCA结合的方法可非常精确的将印文图像进行配准，对后续的特征提取及识别工作提供了很好的支持。进一步，为了将研究成果紧密的结合到真实的应用环境中，本文设计并实现了智能印章识别系统(IntelliSIS)，本文对IntelliSIS系统的设计框架、系统实现、系统性能等作了详细介绍。
As the form of credit system of Chinese society, seal is very important for this. By the development of technology, the producing techniques have a great improvement. With this improvement, the difficulty of fake seal recognition is also being higher and higher and the validity of traditional manual comparison is greatly low. Thus the automatic computer identification research arose. Automatic seal identification belongs to pattern recognition domain and includes seal segmentation, seal registration, feature extraction and feature identification these steps. Recently, some good research results have been applied in practice. But because of the complexity of seal images, there are still many problems need to be resolved.
According to research, we have proposed a new method, applying the Pulse Couple Neural Network (PCNN) in automatic seal identification. By using the auto wave character of PCNN, this method can locate the seal in the image. After we get the location of the seal imprint we can obtain the available critical color information of the seal imprint itself and segment seal texture from the image efficiently by using the accurate threshold. At the same time we also applied the Principal Components Analysis (PCA) neural network in seal registration successfully. The PCA can obtain the approximate direction of the seal imprint by using the global distribution of the image pixels of seal imprint. After this it can use the local search method for precisely registration and get a very good result.
To validate the PCNN and PCA methods in practice, we did the simulation by using the real seal images and PCNN and PCA in MATLAB. Through the experiment, the PCNN method is efficient, it can segment seal texture from the whole seal image precisely and PCA method can also register the seals precisely. This will helpful to the subsequent steps. Furthermore, to integrate the methods in a practical applied environment, we have designed and implemented an Intelligent Seal Identification System (IntelliSIS), it will introduce the details of system framework, implementation and system performance in this article.
引文
[1] Chen Y.S., Automatic Identification for a Chinese Seal Image. Partern Recogonition, vol29, nol 1, 1996, p1807-1820.
[2] Ueda, K., Nakamura, Y., Automatic Verification of Seal-Impresion Patern, Proccedings of 7th Int. Conf. on Patern Recogonition, 1984, p1019-1021.
[3] Fan, T. J., Tsai W. S., Automatic Chinese Seal Identification, Computer Vision, Graphics, and Image Processing, vo 125, 1984, pp311-330.
[4] Chen, C., Tsai, W. S. Automatic identification of Chinese seal with arbitrary or partial shapes. Computer Processing of Chinese & Oriental Languages, 1986, 2(4):p216-233.
[5] Lee S, Kim J.H. Unconstrained seal imprint verification using stroke graph matching. Pattern Recognition, 1989, 22(6):653-664.
[6] 张先萌，罗安玉等，用矩不变量实现印鉴自动识别，电子学报，1995，23(4)：p100-103．
[7] 曾东武，刘重庆等，Fourier描述符的印鉴识别方法，上海交通大学学报，1998，32(9)：p43-45．
[8] 胡庆，杨静宇等，基于知识的印鉴鉴别方法，自动化学报，1991，17(6)：p696-703．
[9] 高文，董胜富等，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)：p338-342．
[10] 廖闻剑，成瑜，银行印鉴的模糊识别方法，南京航空航天大学学报，1998，30(6)：p663-667．
[11] L Li, AJ Walker, Verification and Validation of the SEAL Software Assessment Tool, Proceedings of the 3rd International Software Engineering Standards Symposium (ISESS'97).
[12] Ryoji Haruki, Takahiko Horiuchi, Hiromitsu Yamada and Kazuhiko Yamamoto, Automatic Seall Verification Using Three-Dimensional Reference Seals, Proceedings of the 1996 Patern Recognition (ICPR'96).
[13] Katsuhiko Ueda, Takeshi Mutoho and Kenichi Matsuo, Automatic Seal Verification System for Seal Imprints on Japanese Bank Checks, Proceedings of the 3rd ICDAR, p665-668, 1997.
[14] 杨新军，王肇沂等，基于环投影模板匹配的印鉴鉴别法，光电子激光，2001，12(11)：p1177-1181．
[15] 左奇，史忠科。一种基于直方图评价函数的快速图像分割方法。计算机工程与应用，2003，19：05-06．
[16] 程杰。一种基于直方图的分割方法。华中理工大学学报，1999，27(01)：84-86．
[17] 李弼程，柳葆芳。基于二维直方图的模糊门限分割方法。数据采集与处理，2003，15(03)：324-329．
[18] 柴本成，柴国钟，姜献锋。一种新的自动阈值图像分割方法。机械，2003，30(04)：34-35．
[19] Ku Chin Lin; Fast thresholding computation by searching for zero derivatives of imagebetween class variance. Industrial Electronics Society, 2001. IECON 01. The 27th Annual Conference of the IEEE, 2001, 1:393-397.
[20] Liu Jianzhuang, Li Wenqing, Tian Yupeng; Automatic thresholding of gray level picturesusing two dimension Otsu method. Circuits and Systems, 1991. Conference Proceedings,China., 1991:325-327.
[21] Wong, A.K.C.; Sahoo, P.K.; A gray level threshold selection method based on maximumentropy principle. Systems, Man and Cybernetics, IEEE Transactions on, 1989, 19(04):866-871.
[22] 景晓军，蔡安妮，孙景鳌。一种基于二维最大类间方差的图像分割算法。通信学报，2001， 22(04)：71-76．
[23] 魏宝刚，李向阳，鲁东明等。彩色图像分割研究进展。计算机科学，1999，26(4)：59-62．
[24] 孔健，唐毅。印鉴提取在票据自动处理中的应用研究。微型电脑应用，2002，18(08)：13-15．
[25] 阴炳皓，赵臣，韩晓军。基于改进的HSI空间模型的目标搜索方法。河北工业大学学报，2003，31(01)：06-10．
[26] Sahoo, P.K.; Farag, A.A.; Yeap, Y.. P.; Threshold selection based on HSI histogram modeling. IEEE International Conference on ,Vol. 1, 1992 ,Pages:351-356.
[27] Zhang Yujin, Yao Yurong, He Yun. Color Image Segmentation Based on HIS Model. High Technology Letters, 1998, 14(01):28-31.
[28] Ueda, K. Extraction of signature and seal imprint from bankchecks by using color information. Proceedings of the Third International Conference, 1995, vol.2:665-668.
[29] Otsu. A threshold selection method from gray-level histogram. IEEE Trans on SMC-9, 1979, 9(3):62-66.
[30] H. J, Caulfield, J. M. Kinser, Finding the shortest path in the shortest time using PCNN's, IEEE Trans. on Neural Networks 10 (3) (May 1999) 604-606.
[31] R. Eckhorn, A. Frien, R. Bauer, High frequency (60-90 Hz) oscillations in primary visual cortex of awake monkey, Neuroreport 4 (3) (Mar. 1993) 243-246.
[32] R. Eckhorn, H. J. Reitboeck, M. Arndt, Feature linking via synchronization among distributed assemblies: Simulation of results from cat cortex, Neural Comput. 2 (3) (1990) 293-307.
[33] X.D. Gu, D. H. Yu, L. M. Zhang, Image thinning using pulse coupled neural network, Pattern Recognition Lett. 25 (9) (Jul. 2004) 1075-1084.
[34] J. L. Johnson, M. L. Padgett, PCNN models and applications, IEEE Trans. On Neural Networks 10 (3) (May 1999) 480-498.
[35] J. L. Johnson, D. Ritter, Observation of periodic waves in a pulse-coupled neural network, Opt. Lett. 18 (15) (1993) 1253-1255.
[36] G. Kuntimad, H. S. Ranganath, Perfect image segmentation using pulse coupled neural networks, IEEE Trans. on Neural Networks 10 (3) (May 1999) 591-598.
[37] J.M. Kinser, Foveation by a pulse-coupled neural network, IEEE Trans. On Neural Networks 10 (3) (May 1999) 621-625.
[38] T. Lindblad, J. M. Kinser, Image Processing using Pulse-Coupled Neural Networks, Springer-Verlag, London ISBN 3-540-76264-7(1998).
[39] R. C. Muresan, Pattern recognition using pulse-coupled neural networks and discrete Fourier transforms, Neurocomputing 51 (Apr. 2003) 487-493.
[40] L. Shang, Z. Yi, A class of binary images thinning using two PCNNs, Neurocomputing (2006), doi: 10.1016 /j.neucom.2006.08.006
[41] R.C. Gonzalez, R. E. Woods, Digital Image Processing, (Prentice Hall, 2004).
[42] 张黔，胡庆，杨静宇。统计和结构模式识别方法结合的多特征印鉴真伪鉴别方法，计算机学报，1995，18(3)：190-198
[43] Medioni G G,Nevatia R. Matching Images Using Linear Features[j]. IEEE Trans PAMI, 1984, 1(6):675-685.
[44] 常军民。基于多特征多分类器融合决策的印鉴识别，浙江工业大学硕士学位论文，2005．
[45] Lifeng Shang, Jian Cheng Lv, Zhang Yi. Rigid medical image registration using PCA neural network. Neurocomputing 69 (2006) 1717-1722.
[46] 朱勇，谭铁牛，王蕴红。基于笔迹的身份鉴别[[J]。自动化学报，2001，27(2)：229-234
[47] 李晗，朱虹等。基于奇异值分解和和支持向量机的印鉴真伪鉴别方法[C]。中国电子&测控学会，2003年会
[48] 王洪群，彭嘉雄，强赞霞。基于边缘和纹理特征相结合的快速人脸精确定位方法。计算机工程与应用，2004，7(27)：27-32．
[49] 张鹤，关泽群。基于灰度共生矩阵的图像纹理分析。中国科技论文在线，www.oaoer.edu.cn．
[50] 孙即祥等。模式识别中的特征提取与计算机视觉不变量。国防工业出版社，2001．
[51] MING-KUEI HU. Visual Pattern Recognition by Moment Invariants. IRE Transactions on Information Theory, 1962, pp.179-187.
[52] HONG Z Q. Algebraic feature extraction of image recognition [J]. Pattern Recognition, 1991, 24(3): 211-219
[53] 黄礼平。具有奇异值分解性质的代数[LJ]。数学学报，1997，40(2)：161-166
[2] Ueda, K., Nakamura, Y., Automatic Verification of Seal-Impresion Patern, Proccedings of 7th Int. Conf. on Patern Recogonition, 1984, p1019-1021.
[3] Fan, T. J., Tsai W. S., Automatic Chinese Seal Identification, Computer Vision, Graphics, and Image Processing, vo 125, 1984, pp311-330.
[4] Chen, C., Tsai, W. S. Automatic identification of Chinese seal with arbitrary or partial shapes. Computer Processing of Chinese & Oriental Languages, 1986, 2(4):p216-233.
[5] Lee S, Kim J.H. Unconstrained seal imprint verification using stroke graph matching. Pattern Recognition, 1989, 22(6):653-664.
[6] 张先萌，罗安玉等，用矩不变量实现印鉴自动识别，电子学报，1995，23(4)：p100-103．
[7] 曾东武，刘重庆等，Fourier描述符的印鉴识别方法，上海交通大学学报，1998，32(9)：p43-45．
[8] 胡庆，杨静宇等，基于知识的印鉴鉴别方法，自动化学报，1991，17(6)：p696-703．
[9] 高文，董胜富等，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)：p338-342．
[10] 廖闻剑，成瑜，银行印鉴的模糊识别方法，南京航空航天大学学报，1998，30(6)：p663-667．
[11] L Li, AJ Walker, Verification and Validation of the SEAL Software Assessment Tool, Proceedings of the 3rd International Software Engineering Standards Symposium (ISESS'97).
[12] Ryoji Haruki, Takahiko Horiuchi, Hiromitsu Yamada and Kazuhiko Yamamoto, Automatic Seall Verification Using Three-Dimensional Reference Seals, Proceedings of the 1996 Patern Recognition (ICPR'96).
[13] Katsuhiko Ueda, Takeshi Mutoho and Kenichi Matsuo, Automatic Seal Verification System for Seal Imprints on Japanese Bank Checks, Proceedings of the 3rd ICDAR, p665-668, 1997.
[14] 杨新军，王肇沂等，基于环投影模板匹配的印鉴鉴别法，光电子激光，2001，12(11)：p1177-1181．
[15] 左奇，史忠科。一种基于直方图评价函数的快速图像分割方法。计算机工程与应用，2003，19：05-06．
[16] 程杰。一种基于直方图的分割方法。华中理工大学学报，1999，27(01)：84-86．
[17] 李弼程，柳葆芳。基于二维直方图的模糊门限分割方法。数据采集与处理，2003，15(03)：324-329．
[18] 柴本成，柴国钟，姜献锋。一种新的自动阈值图像分割方法。机械，2003，30(04)：34-35．
[19] Ku Chin Lin; Fast thresholding computation by searching for zero derivatives of imagebetween class variance. Industrial Electronics Society, 2001. IECON 01. The 27th Annual Conference of the IEEE, 2001, 1:393-397.
[20] Liu Jianzhuang, Li Wenqing, Tian Yupeng; Automatic thresholding of gray level picturesusing two dimension Otsu method. Circuits and Systems, 1991. Conference Proceedings,China., 1991:325-327.
[21] Wong, A.K.C.; Sahoo, P.K.; A gray level threshold selection method based on maximumentropy principle. Systems, Man and Cybernetics, IEEE Transactions on, 1989, 19(04):866-871.
[22] 景晓军，蔡安妮，孙景鳌。一种基于二维最大类间方差的图像分割算法。通信学报，2001， 22(04)：71-76．
[23] 魏宝刚，李向阳，鲁东明等。彩色图像分割研究进展。计算机科学，1999，26(4)：59-62．
[24] 孔健，唐毅。印鉴提取在票据自动处理中的应用研究。微型电脑应用，2002，18(08)：13-15．
[25] 阴炳皓，赵臣，韩晓军。基于改进的HSI空间模型的目标搜索方法。河北工业大学学报，2003，31(01)：06-10．
[26] Sahoo, P.K.; Farag, A.A.; Yeap, Y.. P.; Threshold selection based on HSI histogram modeling. IEEE International Conference on ,Vol. 1, 1992 ,Pages:351-356.
[27] Zhang Yujin, Yao Yurong, He Yun. Color Image Segmentation Based on HIS Model. High Technology Letters, 1998, 14(01):28-31.
[28] Ueda, K. Extraction of signature and seal imprint from bankchecks by using color information. Proceedings of the Third International Conference, 1995, vol.2:665-668.
[29] Otsu. A threshold selection method from gray-level histogram. IEEE Trans on SMC-9, 1979, 9(3):62-66.
[30] H. J, Caulfield, J. M. Kinser, Finding the shortest path in the shortest time using PCNN's, IEEE Trans. on Neural Networks 10 (3) (May 1999) 604-606.
[31] R. Eckhorn, A. Frien, R. Bauer, High frequency (60-90 Hz) oscillations in primary visual cortex of awake monkey, Neuroreport 4 (3) (Mar. 1993) 243-246.
[32] R. Eckhorn, H. J. Reitboeck, M. Arndt, Feature linking via synchronization among distributed assemblies: Simulation of results from cat cortex, Neural Comput. 2 (3) (1990) 293-307.
[33] X.D. Gu, D. H. Yu, L. M. Zhang, Image thinning using pulse coupled neural network, Pattern Recognition Lett. 25 (9) (Jul. 2004) 1075-1084.
[34] J. L. Johnson, M. L. Padgett, PCNN models and applications, IEEE Trans. On Neural Networks 10 (3) (May 1999) 480-498.
[35] J. L. Johnson, D. Ritter, Observation of periodic waves in a pulse-coupled neural network, Opt. Lett. 18 (15) (1993) 1253-1255.
[36] G. Kuntimad, H. S. Ranganath, Perfect image segmentation using pulse coupled neural networks, IEEE Trans. on Neural Networks 10 (3) (May 1999) 591-598.
[37] J.M. Kinser, Foveation by a pulse-coupled neural network, IEEE Trans. On Neural Networks 10 (3) (May 1999) 621-625.
[38] T. Lindblad, J. M. Kinser, Image Processing using Pulse-Coupled Neural Networks, Springer-Verlag, London ISBN 3-540-76264-7(1998).
[39] R. C. Muresan, Pattern recognition using pulse-coupled neural networks and discrete Fourier transforms, Neurocomputing 51 (Apr. 2003) 487-493.
[40] L. Shang, Z. Yi, A class of binary images thinning using two PCNNs, Neurocomputing (2006), doi: 10.1016 /j.neucom.2006.08.006
[41] R.C. Gonzalez, R. E. Woods, Digital Image Processing, (Prentice Hall, 2004).
[42] 张黔，胡庆，杨静宇。统计和结构模式识别方法结合的多特征印鉴真伪鉴别方法，计算机学报，1995，18(3)：190-198
[43] Medioni G G,Nevatia R. Matching Images Using Linear Features[j]. IEEE Trans PAMI, 1984, 1(6):675-685.
[44] 常军民。基于多特征多分类器融合决策的印鉴识别，浙江工业大学硕士学位论文，2005．
[45] Lifeng Shang, Jian Cheng Lv, Zhang Yi. Rigid medical image registration using PCA neural network. Neurocomputing 69 (2006) 1717-1722.
[46] 朱勇，谭铁牛，王蕴红。基于笔迹的身份鉴别[[J]。自动化学报，2001，27(2)：229-234
[47] 李晗，朱虹等。基于奇异值分解和和支持向量机的印鉴真伪鉴别方法[C]。中国电子&测控学会，2003年会
[48] 王洪群，彭嘉雄，强赞霞。基于边缘和纹理特征相结合的快速人脸精确定位方法。计算机工程与应用，2004，7(27)：27-32．
[49] 张鹤，关泽群。基于灰度共生矩阵的图像纹理分析。中国科技论文在线，www.oaoer.edu.cn．
[50] 孙即祥等。模式识别中的特征提取与计算机视觉不变量。国防工业出版社，2001．
[51] MING-KUEI HU. Visual Pattern Recognition by Moment Invariants. IRE Transactions on Information Theory, 1962, pp.179-187.
[52] HONG Z Q. Algebraic feature extraction of image recognition [J]. Pattern Recognition, 1991, 24(3): 211-219
[53] 黄礼平。具有奇异值分解性质的代数[LJ]。数学学报，1997，40(2)：161-166