汽车牌照识别的初步研究
|
摘要
智能交通管理系统是近年来国际上最引人注目、研究发展速度最快的研究领域之一，智能交通管理系统是21世纪世界道路交通的发展趋势。高速公路的不断发展和车辆管理体制的不断完善，使图像场景日益简单化和标准化，这为以图像理解为基础的智能交通管理系统进入实际应用领域提供了契机。汽车牌照自动识别系统正是在这种应用背景下研制出来的能够自动、实时地检测车辆经过和识别汽车牌照的一种智能交通管理系统。而车牌自动识别技术在国际上尚处于研究阶段，已能在实验室实现，并且效果良好，但还缺少实际应用的考验，有待进一步的研究和应用。
一个典型的汽车牌照识别系统包括汽车牌照图像的采集，车牌字符的提取和车牌字符识别三大部分。车牌字符提取是汽车牌照识别系统中的难点，本论文采用了多层次分割方法，有效的解决了复杂环境下的汽车牌照定位和提取。对车牌数字识别采用了投影—模板匹配法实现了对车牌数字识别，该方法具有实时性好的特点。
Mainly used in highway traffic management, License Plate Recognition (LPR) System is one kind of Intelligent Transportation Systems (ITS). LPR finds its increasingly wide and important applications in highway electronic toll collection, red-light violation enforcement, secure-access control at parking lots and locating vehicles stolen, or those registered to fugitives, criminals and smugglers. A typical LPR System consists of a video image-acquisition subsystem, an image-preprocessing unit, a plate character isolation & recognition engine and a local storage or remote transmission subsystem.
ITS and LPR are reviewed briefly in the first chapter of this paper. To give a whole idea of LPR, Chapter Two describes how a LPR System works. A multiplayer method is used in division of the License Plate to solve how to go to and pick up the plate character. A method of the projection and correlation method is used in identifiving the plate character.
Two important subsystems of LPR are considered in Chapter Three and Chapter Four separately: Plate Character Isolation & Recognition. Plate Character Isolation refers to the process to find the plate in a car image and separate each plate characters, which include Chinese, English and digital characters. Just as its name implies, Plate Character Recognition means recognizing the isolated plate characters and in turn identifying the whole license plate.
一个典型的汽车牌照识别系统包括汽车牌照图像的采集，车牌字符的提取和车牌字符识别三大部分。车牌字符提取是汽车牌照识别系统中的难点，本论文采用了多层次分割方法，有效的解决了复杂环境下的汽车牌照定位和提取。对车牌数字识别采用了投影—模板匹配法实现了对车牌数字识别，该方法具有实时性好的特点。
Mainly used in highway traffic management, License Plate Recognition (LPR) System is one kind of Intelligent Transportation Systems (ITS). LPR finds its increasingly wide and important applications in highway electronic toll collection, red-light violation enforcement, secure-access control at parking lots and locating vehicles stolen, or those registered to fugitives, criminals and smugglers. A typical LPR System consists of a video image-acquisition subsystem, an image-preprocessing unit, a plate character isolation & recognition engine and a local storage or remote transmission subsystem.
ITS and LPR are reviewed briefly in the first chapter of this paper. To give a whole idea of LPR, Chapter Two describes how a LPR System works. A multiplayer method is used in division of the License Plate to solve how to go to and pick up the plate character. A method of the projection and correlation method is used in identifiving the plate character.
Two important subsystems of LPR are considered in Chapter Three and Chapter Four separately: Plate Character Isolation & Recognition. Plate Character Isolation refers to the process to find the plate in a car image and separate each plate characters, which include Chinese, English and digital characters. Just as its name implies, Plate Character Recognition means recognizing the isolated plate characters and in turn identifying the whole license plate.
引文
[1] 边肇祺，张学工等，模式识别，清华大学出版社。
[2] 章毓晋，图像分割，科学出版社。
[3] 刘枫，静态图像压缩编码研究，成都理工大学硕士论文。
[4] 崔屹，图像处理与分析——数学形态学方法及应用，科学出版社。
[5] 袁志伟，潘晓露，陈艾，李一民，车辆牌照定位的算法研究，昆明理工大学学报[J]，2001，26(2)56～40
[6] 王年等，基于神经网络的汽车牌照自动识别，安徽大学学报[J]，2000，24(1)，45～50
[7] 李介谷，蔡国廉。计算机模式识别技术M1．上海：上海交通大学出版社，1992。
[8] 赵明生，包丛笑 车牌识别技术在交通工程和城市安全中的作用及应用前景．见交通系统发展趋势国际学术研讨会论文集 北京：人民交通出版社，1997．
[9] 寥全局，宣国荣．车辆牌照的自动分割J1．微型电脑应用，1999(7)：10
[10] 黎洪松．数字视频技术及其应用LM工北京：清华大学出版社，1998．
[11] Cast eman KR．数字图像处理LM；．北京：北京工业出版社，1998．
[12] 局锐等、复杂背景中目标图像的提取与跟踪LJ：．系统工程与电子技术．1997(g)．
[13] 荆仁杰等 计算机图像处理 浙江大学出版社
[14] 李介谷等 计算机模式识别技术 上海交通大学出版社
[15] 段里仁．智能交通系统在我国道路交通管理中的应用．’97北京智能交通系统发展趋势国际学术研讨会论文集：p 15-22．
[16] 刘济林，马洪庆，叶秀清．汽车牌照自动识别系统．’97北京智能交通系统发展趋势国际学术研讨会论文集：p 155-159．
[17] R.M. Haralick, S.R. Steraberg, X. Zhuang. Image Analysis Using Mathematical Morphology. IEEE Transaction on Pattern Analysis and Machine Intelligent, Vol. 2529, No. 4, July 1987.
[18] Chung-Mu. Hwang, et al. A PC Based Car Licence Plate Reader. SPIE Vol.1823(1992)．
[19]荆仁杰，叶秀清，徐胜荣，陈存椿．计算机图像处理．浙江大学出版社．
[20]Duda R O, Hart P E. Use of Hough Transformation to Detect Lines and Curves in Picture. Comm. ACM 15(1):11-15, 1972.
[21]Stephen W. Lan, Aunradha Bhate and Sargue N. Srihari. A Sliding Window Technique for Word Recognition. SPIE Vol. 2422.
[22]李立源，陈维南．一种强鲁棒的完全确定型的快速阈值化方法．PR&AI．Vo1．6，No．3，Sep．1993．
[23]张引，潘云鹤，彩色汽车图像牌照定位新方法。中国图像图形学报[J]。2001，6(4)：374～377
[24]赵学春，戚飞虎，基于彩色分割的车牌自动识别技术，上海交通大学学报[J]，1998-32 (10)；4-9
[25]Haralick R.M.. Digtal Step Edges from Zero Crossing of Second Directional and Machine Intelligence[J] 699～714
[26]Raus M,Kreft L.Reading car license plates by the use of artificial neural networks.In:Proceedings of the 1995 IEEE 38th Midwest Symposium on Circuits andSystems[J],NJ,USA:IEEE.1995,Part 1(of 2):538～541
[27]Nalwa V.S..On Detecting Edges.IEEE Trans on Pattern Analysis and Machine Intelligence[J].699-714
[28]Haraliek R.M.. Digtal Step Edges from Zero Crossing of Second Directional and Machine Intelligence [J]699～714
[29]Buster T.C..A Critical View of Pyramid Segmentation Algorithms.Pattern Recognition Letters[J].605-617
[30]Buster T.C..A Critical View of Pyramid Segmentation Algorithms. Pattern Recognition Letters[J].605-617
[31]Chow C K, Kaneko T. Automatic Boundary Detection of the Left Ventricle from Cineangiograns. Comput. Bioned. Res, 5, 388-410, 1972.
[32]T. Pun. Entropic Thresholding, A New Approach. CGIP Vol. 16, p210-239, 1981.
[33]Bir Bhanu. Adaptive Image Segmentation Using a Genetic Algorithm. IEEE. Transactions on systems, man, and cybernetics, Vol. 25, p1543-1567, No. 12, Dec., 1995.
[34]J. Mamtas. An Overview of Character recognition methodologies. Pattern Recognition, Vol. 19, No. 6, 1986.
[35]张炘中．计算机汉字自动识别技术．中国印刷，26期，1989．
[36]吴佑寿，丁晓青，汉字识别一原理、方法和实现，高等教育出版社．
[37]Kamal Jambi, Rafaat Shamsan. An Experimental Approach for Recognizing Handwritten Arabic Words. ACCV' 95 Second Asian Conference on Computer Vision.
[38]Jin Wang and Jack Jean, Segmentation of Merged Characters By Neural Networks and Shortest Path. Pattern Recognition, Vol. 27, No. 5, pp. 649-658, 1994.
[39]Seong-Whan Lee, Jeong-Seon Park, Nonlinear Shape Normalization Methods For The Recognition of Large-Set Handwritten Character. Pattern Recognition, Vol. 27, No. 7, pp. 895-902, 1994.
[40]洪沁，何振亚，手写体数字的神经网络识别方法，PR&AI，Vol．7 No．1 March 1994．
[41]张中，汉字识别技术，清华大学出版社，广西科学技术出版社．
[42]Hildebrandt, T.H. and Liu, W., Optical Recognition of Handwritten Chinese Characters: Advances Since 1980, Pattern Recognition 26 (2) (1993) 205-225.
[43]Mori, S., Yamamoto, K. and Yasuda, M., Research on Machine Recognition of Handprinted Characters, IEEE Transactions on Pattern Analysis and Machine Intelligece 6(4) (1984) 386-405.
[44]Sue, C. Y., Berthod, M. and Mori, S., Automatic Recognition of Handprinted Characters-The State of the Art, Proceedings of the IEEE 68(4) (1980) 469-487.
[45]Yamamoto, K. and Rosenfeld, A., Recognition of Handprinted Kanji Characters by a Relaxation Method, Proc. 6th Int. Conf. Pattern Recognition (1982) 395-398
[46]戴汝为，语义句法模式识别方法极其应用，模式识别与人工智能89(2)(1995)90-93．
[47]王润生．图像理解．国防科技大学出版社，1998
[48]吴一全，未兆达．图像处理中阀值选取方法30年(1962-1992)的进展(一)．数据采集与处理，1993；8(3)：193—201
[49]吴一全，未兆达．图像处理中阈值选取方法30年(1962-1992)的进展(二)．数据采集与处理，1993；8(4)：268-281
[50]I.T. Jolliffe. Principal Component Analysis. New York: Springer-Verlag, 1986.
[51]K. Pearson, On Lines and Planes of Closest Fit to Systems of Points in Space, Phil. Mag. 6th series 1901.
[52]H. Hotelling, Analysis of a Complex of Statistical Variables into Principal Components, J. Educational Psychol., September 1993.
[53]S. Watanabe, Karhunen-Loeve Expansion and Factor Analysis - Theoretical Remarks and Applications, Tran. Of the 4th Prague Conf. On Information Theory. Statistical Decision Functions, Random Processes, p. 635, 1965.
[54]E. Oja. Principal Component Analysis by Homogeneous Neural Networks, Part Ⅰ:The Weighted Subspace Criterion. IEICE Trans. Inf. & Sys., 1992; E75-D (3):366-375
[55]E. Oja. Principal Component Analysis by Homogeneous Neural Networks, Part Ⅱ, Analysis and Extensions of the Learning Algorithm, IEICE Trans. Inf. & Sys., 1992;E75-D (3): 376-382
[56]Hee-Seon Paark, Sang-Yup Kim and Seong-Whan Lee, Gray-Scale Handwritten Character Recognition Based on Princial Features, SPIE Vol. 3027, 40-49
[57]C. Shang, K. Brown. Principal Features-Based Texture Classification with Neural Networks. Pattern Recognition, Vol. 27, No. 5, p. 675-687, 1994.
[58]Bernhard Scholk(?)pf, Alexander Smola, Klaus-Robert MUller, Nonlinear Component Analysis as a Kernel Eigenvalue Problem, Letter, 1299-1319.
[2] 章毓晋，图像分割，科学出版社。
[3] 刘枫，静态图像压缩编码研究，成都理工大学硕士论文。
[4] 崔屹，图像处理与分析——数学形态学方法及应用，科学出版社。
[5] 袁志伟，潘晓露，陈艾，李一民，车辆牌照定位的算法研究，昆明理工大学学报[J]，2001，26(2)56～40
[6] 王年等，基于神经网络的汽车牌照自动识别，安徽大学学报[J]，2000，24(1)，45～50
[7] 李介谷，蔡国廉。计算机模式识别技术M1．上海：上海交通大学出版社，1992。
[8] 赵明生，包丛笑 车牌识别技术在交通工程和城市安全中的作用及应用前景．见交通系统发展趋势国际学术研讨会论文集 北京：人民交通出版社，1997．
[9] 寥全局，宣国荣．车辆牌照的自动分割J1．微型电脑应用，1999(7)：10
[10] 黎洪松．数字视频技术及其应用LM工北京：清华大学出版社，1998．
[11] Cast eman KR．数字图像处理LM；．北京：北京工业出版社，1998．
[12] 局锐等、复杂背景中目标图像的提取与跟踪LJ：．系统工程与电子技术．1997(g)．
[13] 荆仁杰等 计算机图像处理 浙江大学出版社
[14] 李介谷等 计算机模式识别技术 上海交通大学出版社
[15] 段里仁．智能交通系统在我国道路交通管理中的应用．’97北京智能交通系统发展趋势国际学术研讨会论文集：p 15-22．
[16] 刘济林，马洪庆，叶秀清．汽车牌照自动识别系统．’97北京智能交通系统发展趋势国际学术研讨会论文集：p 155-159．
[17] R.M. Haralick, S.R. Steraberg, X. Zhuang. Image Analysis Using Mathematical Morphology. IEEE Transaction on Pattern Analysis and Machine Intelligent, Vol. 2529, No. 4, July 1987.
[18] Chung-Mu. Hwang, et al. A PC Based Car Licence Plate Reader. SPIE Vol.1823(1992)．
[19]荆仁杰，叶秀清，徐胜荣，陈存椿．计算机图像处理．浙江大学出版社．
[20]Duda R O, Hart P E. Use of Hough Transformation to Detect Lines and Curves in Picture. Comm. ACM 15(1):11-15, 1972.
[21]Stephen W. Lan, Aunradha Bhate and Sargue N. Srihari. A Sliding Window Technique for Word Recognition. SPIE Vol. 2422.
[22]李立源，陈维南．一种强鲁棒的完全确定型的快速阈值化方法．PR&AI．Vo1．6，No．3，Sep．1993．
[23]张引，潘云鹤，彩色汽车图像牌照定位新方法。中国图像图形学报[J]。2001，6(4)：374～377
[24]赵学春，戚飞虎，基于彩色分割的车牌自动识别技术，上海交通大学学报[J]，1998-32 (10)；4-9
[25]Haralick R.M.. Digtal Step Edges from Zero Crossing of Second Directional and Machine Intelligence[J] 699～714
[26]Raus M,Kreft L.Reading car license plates by the use of artificial neural networks.In:Proceedings of the 1995 IEEE 38th Midwest Symposium on Circuits andSystems[J],NJ,USA:IEEE.1995,Part 1(of 2):538～541
[27]Nalwa V.S..On Detecting Edges.IEEE Trans on Pattern Analysis and Machine Intelligence[J].699-714
[28]Haraliek R.M.. Digtal Step Edges from Zero Crossing of Second Directional and Machine Intelligence [J]699～714
[29]Buster T.C..A Critical View of Pyramid Segmentation Algorithms.Pattern Recognition Letters[J].605-617
[30]Buster T.C..A Critical View of Pyramid Segmentation Algorithms. Pattern Recognition Letters[J].605-617
[31]Chow C K, Kaneko T. Automatic Boundary Detection of the Left Ventricle from Cineangiograns. Comput. Bioned. Res, 5, 388-410, 1972.
[32]T. Pun. Entropic Thresholding, A New Approach. CGIP Vol. 16, p210-239, 1981.
[33]Bir Bhanu. Adaptive Image Segmentation Using a Genetic Algorithm. IEEE. Transactions on systems, man, and cybernetics, Vol. 25, p1543-1567, No. 12, Dec., 1995.
[34]J. Mamtas. An Overview of Character recognition methodologies. Pattern Recognition, Vol. 19, No. 6, 1986.
[35]张炘中．计算机汉字自动识别技术．中国印刷，26期，1989．
[36]吴佑寿，丁晓青，汉字识别一原理、方法和实现，高等教育出版社．
[37]Kamal Jambi, Rafaat Shamsan. An Experimental Approach for Recognizing Handwritten Arabic Words. ACCV' 95 Second Asian Conference on Computer Vision.
[38]Jin Wang and Jack Jean, Segmentation of Merged Characters By Neural Networks and Shortest Path. Pattern Recognition, Vol. 27, No. 5, pp. 649-658, 1994.
[39]Seong-Whan Lee, Jeong-Seon Park, Nonlinear Shape Normalization Methods For The Recognition of Large-Set Handwritten Character. Pattern Recognition, Vol. 27, No. 7, pp. 895-902, 1994.
[40]洪沁，何振亚，手写体数字的神经网络识别方法，PR&AI，Vol．7 No．1 March 1994．
[41]张中，汉字识别技术，清华大学出版社，广西科学技术出版社．
[42]Hildebrandt, T.H. and Liu, W., Optical Recognition of Handwritten Chinese Characters: Advances Since 1980, Pattern Recognition 26 (2) (1993) 205-225.
[43]Mori, S., Yamamoto, K. and Yasuda, M., Research on Machine Recognition of Handprinted Characters, IEEE Transactions on Pattern Analysis and Machine Intelligece 6(4) (1984) 386-405.
[44]Sue, C. Y., Berthod, M. and Mori, S., Automatic Recognition of Handprinted Characters-The State of the Art, Proceedings of the IEEE 68(4) (1980) 469-487.
[45]Yamamoto, K. and Rosenfeld, A., Recognition of Handprinted Kanji Characters by a Relaxation Method, Proc. 6th Int. Conf. Pattern Recognition (1982) 395-398
[46]戴汝为，语义句法模式识别方法极其应用，模式识别与人工智能89(2)(1995)90-93．
[47]王润生．图像理解．国防科技大学出版社，1998
[48]吴一全，未兆达．图像处理中阀值选取方法30年(1962-1992)的进展(一)．数据采集与处理，1993；8(3)：193—201
[49]吴一全，未兆达．图像处理中阈值选取方法30年(1962-1992)的进展(二)．数据采集与处理，1993；8(4)：268-281
[50]I.T. Jolliffe. Principal Component Analysis. New York: Springer-Verlag, 1986.
[51]K. Pearson, On Lines and Planes of Closest Fit to Systems of Points in Space, Phil. Mag. 6th series 1901.
[52]H. Hotelling, Analysis of a Complex of Statistical Variables into Principal Components, J. Educational Psychol., September 1993.
[53]S. Watanabe, Karhunen-Loeve Expansion and Factor Analysis - Theoretical Remarks and Applications, Tran. Of the 4th Prague Conf. On Information Theory. Statistical Decision Functions, Random Processes, p. 635, 1965.
[54]E. Oja. Principal Component Analysis by Homogeneous Neural Networks, Part Ⅰ:The Weighted Subspace Criterion. IEICE Trans. Inf. & Sys., 1992; E75-D (3):366-375
[55]E. Oja. Principal Component Analysis by Homogeneous Neural Networks, Part Ⅱ, Analysis and Extensions of the Learning Algorithm, IEICE Trans. Inf. & Sys., 1992;E75-D (3): 376-382
[56]Hee-Seon Paark, Sang-Yup Kim and Seong-Whan Lee, Gray-Scale Handwritten Character Recognition Based on Princial Features, SPIE Vol. 3027, 40-49
[57]C. Shang, K. Brown. Principal Features-Based Texture Classification with Neural Networks. Pattern Recognition, Vol. 27, No. 5, p. 675-687, 1994.
[58]Bernhard Scholk(?)pf, Alexander Smola, Klaus-Robert MUller, Nonlinear Component Analysis as a Kernel Eigenvalue Problem, Letter, 1299-1319.