汽车发动机号计算机辅助甄别系统的研究
|
摘要
机动车辆出厂时,在发动机和车架上均要打上识别号,这两个号码统称为车辆鉴别码。在车辆管理部门登记时,要对其进行拓印并保存,这时的拓片我们称之为原始拓片。在日后的车辆年审、过户、改装等业务时,均要对发动机号和车架号再次进行拓印,并与车辆注册登记时的原始拓片进行比对,以排除非法偷换发动机或车架等违法行为。车辆鉴别代号相当于车辆的“身份证”,甄别发动机号和车架号的非自然改动情况,是车辆年审时判断车辆合法性的重要依据。
随着我国经济的发展,机动车数量急剧增长,国内大中型城市机动车平均保有量已达50万辆以上。如此庞大的数量,全都依靠技术人员手工和经验来鉴别;加之近年来盗抢机动车的犯罪活动不断增多,给公安交通管理部门的工作带来极大的难度。使用计算机辅助识别,提高工作效率和精度已成为交通管理部门、乃至全社会的迫切需求。
本文将图像处理技术应用于车辆鉴别号码的甄别,它主要由图像预处理和图像比对两部分组成。对图像的预处理,主要通过图像数字化、边缘检测、去除噪声、定位目标、字符分割等步骤,建立发动机号和车架号的样本信息,包括:拓片的图像数据(原始拓片样本图像和待检测拓片的比对图像)、以及这些图像中字符对象的坐标数据。
在图像预处理中,噪声的去除程度对后面的比对效果有极大的影响,车辆鉴别码拓片图像中噪声呈现连续性、块状性的特点,使用综合方法去噪是本文研究的重点之一;结合拓片字符的几何特征,本文使用基于模板匹配的方法,通过两次分割将单个字符分割出来,有效克服字符的粘连和断裂;比对算法将原始拓片图像和待检测拓片的比对图像进行图像配准,生成比对信息,比对信息包括两幅图像中所有字符对象的坐标信息和对应区域的重合程度。
比对系统的使用可以缩短交通管理部门的业务处理时间,提高工作效率;可以为发动机号和车架号的甄别提供一个直观和科学的依据,其具有很大的实用价值。
Vehicle identification number(VlN) will be put on engine and frame as a seal when vehicles leave factory. A new vehicle's numbers must be stamped and the rubbings, which is called original label, must be perserved in the Traffic Department. The numbers will be stamped again and compared with the original label in the operations for Annual vehicle check-up, refit or transfer etc., and it will protect against that the numbers have been illegal changed. VIN is one and only ID of vehicles. Discriminating the numbers which have been factitious changed is important and it is the key to identify the validity of vehicles.
With the rapid development of economy in our country, the quantity of vehicle become more and more. There are more than 500 thousand vehicles in the medium and large cities of our country on average. It's hard to discriminate such a large number of rubbings by human work. And with the increasing of criminal activities, it becomes more difficult to identify the validity of vehicles for Traffic Department, so it's necessary to improve efficiency and precision by using computers.
In this thesis, a technique of image processing and its application in VIN is demonstrated. The two important parts of the technique are the image preprocessing algorithm and the image comparison algorithm. The image preprocessing algorithm is used in the register of vehicles, which includes Image Digitization, edge detection, de-noising, characters segmentation. This will set up data of the image, which includes the coordinate information of all the characters in the image.
Image de-noising will greatly influence the latter compared results, so it is one of the emphases of study; On Character segmentation, template matching method is improved to overcome characters' conglutination and rupture effectively; The image comparison algorithm is used in image registration and to test the validity of vehicle through the similarity degree between the relevant characters.
The use of comparison system will minimize the time of affair transaction and increase the efficiency for the traffic apartment; It will provided direct basis for the validity of vehicle, and has a great value of practicality.
随着我国经济的发展,机动车数量急剧增长,国内大中型城市机动车平均保有量已达50万辆以上。如此庞大的数量,全都依靠技术人员手工和经验来鉴别;加之近年来盗抢机动车的犯罪活动不断增多,给公安交通管理部门的工作带来极大的难度。使用计算机辅助识别,提高工作效率和精度已成为交通管理部门、乃至全社会的迫切需求。
本文将图像处理技术应用于车辆鉴别号码的甄别,它主要由图像预处理和图像比对两部分组成。对图像的预处理,主要通过图像数字化、边缘检测、去除噪声、定位目标、字符分割等步骤,建立发动机号和车架号的样本信息,包括:拓片的图像数据(原始拓片样本图像和待检测拓片的比对图像)、以及这些图像中字符对象的坐标数据。
在图像预处理中,噪声的去除程度对后面的比对效果有极大的影响,车辆鉴别码拓片图像中噪声呈现连续性、块状性的特点,使用综合方法去噪是本文研究的重点之一;结合拓片字符的几何特征,本文使用基于模板匹配的方法,通过两次分割将单个字符分割出来,有效克服字符的粘连和断裂;比对算法将原始拓片图像和待检测拓片的比对图像进行图像配准,生成比对信息,比对信息包括两幅图像中所有字符对象的坐标信息和对应区域的重合程度。
比对系统的使用可以缩短交通管理部门的业务处理时间,提高工作效率;可以为发动机号和车架号的甄别提供一个直观和科学的依据,其具有很大的实用价值。
Vehicle identification number(VlN) will be put on engine and frame as a seal when vehicles leave factory. A new vehicle's numbers must be stamped and the rubbings, which is called original label, must be perserved in the Traffic Department. The numbers will be stamped again and compared with the original label in the operations for Annual vehicle check-up, refit or transfer etc., and it will protect against that the numbers have been illegal changed. VIN is one and only ID of vehicles. Discriminating the numbers which have been factitious changed is important and it is the key to identify the validity of vehicles.
With the rapid development of economy in our country, the quantity of vehicle become more and more. There are more than 500 thousand vehicles in the medium and large cities of our country on average. It's hard to discriminate such a large number of rubbings by human work. And with the increasing of criminal activities, it becomes more difficult to identify the validity of vehicles for Traffic Department, so it's necessary to improve efficiency and precision by using computers.
In this thesis, a technique of image processing and its application in VIN is demonstrated. The two important parts of the technique are the image preprocessing algorithm and the image comparison algorithm. The image preprocessing algorithm is used in the register of vehicles, which includes Image Digitization, edge detection, de-noising, characters segmentation. This will set up data of the image, which includes the coordinate information of all the characters in the image.
Image de-noising will greatly influence the latter compared results, so it is one of the emphases of study; On Character segmentation, template matching method is improved to overcome characters' conglutination and rupture effectively; The image comparison algorithm is used in image registration and to test the validity of vehicle through the similarity degree between the relevant characters.
The use of comparison system will minimize the time of affair transaction and increase the efficiency for the traffic apartment; It will provided direct basis for the validity of vehicle, and has a great value of practicality.
引文
[1]陈焕昌.汽车发动机号码车架号码被改动的检验方法[J].广东公安科技,2006.2
[2]陈蕊丽.复现机动车标识号码的检验原理及其方法[J].中国人民公安大学学报(自然科学版).2005,No.2
[3]史忠科,曹力主编.交通图像检测与分析[M].北京:科学出版社,2007
[4]袁满荣。车架号计算机辅助识别系统的研制[D].昆明理工大学,2005
[5]Rafael C.Gonzalez著.阮秋琦,阮宇智等译.数字图像处理(第二版)[M].电子工业出版社
[6]何东健主编.数字图像处理[M].西安电子科技大学出版社
[7]章敏晋.图像分割[M].科学出版社.2001
[8]Sousa J B.An integrated development process for automated or driver-assist vehicles[J].Proceedings of SPIE,2002,47(15):228-237
[9]P.Soille,Morphological Image Analysis:Principles and Applications[M],Berlin,Germany:Springer-Verlag,1999
[10]D.Van De Ville etal.Noise reduction by fuzzy image filtering[J],IEEE Trans.Fuzzy Syst,vol.11,pp.429-436,Aug.2003.
[11]K.Suzuki,I.Horiba,and N.Sugie.Neural edge enhancer for supervised edge enhancement from noisy images[J],IEEE Trans,Pattern Anal.Machine Intell,vol.25,pp.I582-1596,Dec.2003
[12]David Beymer,Philip McLauchlan.A Real-time computer vision system for vehicle tracking and traffic surveillance[J].Transportation Research:Part C,1998,6(4):271-288
[13]Comelli P,Ferragina P,Nottumo M.Optical recognition of motor vehicle number plate[J].IEEE Transactions On Vehicular Technology,1995,44(4):790-799
[14]张引,潘云鹤,彩色汽车图像牌照定位新方法[J],中国图像图形学报,Vo.61,No.4.AP.r2001,PE374-377.
[15]赵雪春,戚飞虎,基于彩色分割的车牌自动识别技术[J],上海交通大学学报,Vol32,No.10,Oct.1998,pp.4-9.
[16]刘奕陈学俭蒋治华,一种利用颜色信息的车牌字符分割新算法[J],计算机应用与软件,Vol.22,No.8,Aug.2005,PP.98-10
[17]陈寅鹏,丁晓青.复杂车辆图像中的车牌定位与字符分割方法[J].红外与激光工程.2004,Vol.33No.1
[18]王宸昊.复杂背景下车牌定位分割技术的研究[D].湘潭大学硕士学位论文.2006
[19]S.U.Lee and S.Y.Chung,A comparative performance study of several global thresholding techniques for segmentation[J],Comput.Vis,Graph,Image Process,vol.52,pp.171-190,1990.
[20]Otsu,Minimum error thresholding[J],Pattern Recognit,vol,19,no.1,pp.41-47,1986.
[21]Y.Zhang and C.Zhang,A new algorithm for character segmentation of license plate[J],in Proc.IEEE Interlligent Vehicles Symp,June 2003,pp.106-109
[22]FengYang,ZhengMa,MeiXie.A Novel APProach for License Plate Character Segmentation[J].Industrial Eleetronies and APPlications,2006,5:1-6
[23]D.-S.Kim and S-L Chien,Automatic car license plate extraction using modified generalized symmetry transform and image warping[J],in Proc.IEEE Int.Symp.Industrial Electronics(ISIE'01),vol.3,June 2001,pp.2022-2027
[24]王瑛楠.车牌自动识别系统的研制与开发[D].西北工业学硕士学位论文.2007
[25]何志伟.基于模型的目标提取及其在智能交通中的应用[D].浙江大学博士论文.2006,4
[26]石振刚,高立群,葛雯.一种基于模糊逻辑的图像自适应去噪算法.东北大学学报(自然科学版),2007,第28卷第6期
[27]左奇,史忠科.基于模糊理论的图像分割方法[D].西北工业大学学报,2003,21(3):313-316
[28]左奇,史忠科.一种基于数学形态学的实时车牌图像分割方法[J].中国图像图形学报,2003,8(3):281-285
[29]杨淑莹.VC++图像处理程序设计[M].清华大学出版社,北方交通大学出版社.2003:158-159.
[30]王良红,冷建华.汽车倾斜牌照中字符的定位与提取[J].电讯技术,2003,4:61.
[31]刘志勇,刘迎建.车牌识别中的图像提取及分割[J].中文信息学报.2000.4
[32]王慧燕.图像边缘检测和图像匹配研究及应用[D].浙江大学博士论文2003,8
[33]刘宁.指纹匹配算法研究[D].山东大学硕士论文2006.4.
[34]龙世渚,黄载禄,曹汉房.一种立体视觉图像的特征选择与匹配方法[J].华中理工大学学报,1994,22(11):28-31
[35]舒丽霞,周成平,彭晓明,丁明跃.基于Hausdorff距离图像配准方法研究[J].中国图像图形学报.2003.12.第8卷(A版)第12期
[36]李震,范湘涛,施建成.基于分窗口相关的遥感图像配准方法[J].中国图像图形学报.2001.02,第6卷(A版)第2期
[37]李秀秀,郑江滨,张艳宁.一种新的自动图像配准技术[J].计算机应用研究.2008.1.第25卷第1期
[38]陈世伟,李世平,管京周,熊楠.一种改进的基于轮廓的图像配准算法[J].计量技术.2005.9.
[39]荣江,王文杰,陈建华.智能车牌识别系统设计与实现[J].微型机与应用.2005.01
[40]管业鹏,童林夙.二维图像特征点快速提取算法[J].中国图像图形学报.2002.12.第7卷(A版)第12期
[41]郑宇,肖南峰.一种基于相似度的汽车车牌号码自动识别系统[J].交通与计算机.2004.第5期,第22卷
[42]杨威,赵剡,许东.基于人眼视觉的结构相似度图像质量评价方法[J].北京航空航天大学学报.2008.第34卷第1期
[43]张健,毛二可,赵保军,王靖.真假目标红外图像相似度检测算法[J].光电工程.2007.第34卷第9期
[2]陈蕊丽.复现机动车标识号码的检验原理及其方法[J].中国人民公安大学学报(自然科学版).2005,No.2
[3]史忠科,曹力主编.交通图像检测与分析[M].北京:科学出版社,2007
[4]袁满荣。车架号计算机辅助识别系统的研制[D].昆明理工大学,2005
[5]Rafael C.Gonzalez著.阮秋琦,阮宇智等译.数字图像处理(第二版)[M].电子工业出版社
[6]何东健主编.数字图像处理[M].西安电子科技大学出版社
[7]章敏晋.图像分割[M].科学出版社.2001
[8]Sousa J B.An integrated development process for automated or driver-assist vehicles[J].Proceedings of SPIE,2002,47(15):228-237
[9]P.Soille,Morphological Image Analysis:Principles and Applications[M],Berlin,Germany:Springer-Verlag,1999
[10]D.Van De Ville etal.Noise reduction by fuzzy image filtering[J],IEEE Trans.Fuzzy Syst,vol.11,pp.429-436,Aug.2003.
[11]K.Suzuki,I.Horiba,and N.Sugie.Neural edge enhancer for supervised edge enhancement from noisy images[J],IEEE Trans,Pattern Anal.Machine Intell,vol.25,pp.I582-1596,Dec.2003
[12]David Beymer,Philip McLauchlan.A Real-time computer vision system for vehicle tracking and traffic surveillance[J].Transportation Research:Part C,1998,6(4):271-288
[13]Comelli P,Ferragina P,Nottumo M.Optical recognition of motor vehicle number plate[J].IEEE Transactions On Vehicular Technology,1995,44(4):790-799
[14]张引,潘云鹤,彩色汽车图像牌照定位新方法[J],中国图像图形学报,Vo.61,No.4.AP.r2001,PE374-377.
[15]赵雪春,戚飞虎,基于彩色分割的车牌自动识别技术[J],上海交通大学学报,Vol32,No.10,Oct.1998,pp.4-9.
[16]刘奕陈学俭蒋治华,一种利用颜色信息的车牌字符分割新算法[J],计算机应用与软件,Vol.22,No.8,Aug.2005,PP.98-10
[17]陈寅鹏,丁晓青.复杂车辆图像中的车牌定位与字符分割方法[J].红外与激光工程.2004,Vol.33No.1
[18]王宸昊.复杂背景下车牌定位分割技术的研究[D].湘潭大学硕士学位论文.2006
[19]S.U.Lee and S.Y.Chung,A comparative performance study of several global thresholding techniques for segmentation[J],Comput.Vis,Graph,Image Process,vol.52,pp.171-190,1990.
[20]Otsu,Minimum error thresholding[J],Pattern Recognit,vol,19,no.1,pp.41-47,1986.
[21]Y.Zhang and C.Zhang,A new algorithm for character segmentation of license plate[J],in Proc.IEEE Interlligent Vehicles Symp,June 2003,pp.106-109
[22]FengYang,ZhengMa,MeiXie.A Novel APProach for License Plate Character Segmentation[J].Industrial Eleetronies and APPlications,2006,5:1-6
[23]D.-S.Kim and S-L Chien,Automatic car license plate extraction using modified generalized symmetry transform and image warping[J],in Proc.IEEE Int.Symp.Industrial Electronics(ISIE'01),vol.3,June 2001,pp.2022-2027
[24]王瑛楠.车牌自动识别系统的研制与开发[D].西北工业学硕士学位论文.2007
[25]何志伟.基于模型的目标提取及其在智能交通中的应用[D].浙江大学博士论文.2006,4
[26]石振刚,高立群,葛雯.一种基于模糊逻辑的图像自适应去噪算法.东北大学学报(自然科学版),2007,第28卷第6期
[27]左奇,史忠科.基于模糊理论的图像分割方法[D].西北工业大学学报,2003,21(3):313-316
[28]左奇,史忠科.一种基于数学形态学的实时车牌图像分割方法[J].中国图像图形学报,2003,8(3):281-285
[29]杨淑莹.VC++图像处理程序设计[M].清华大学出版社,北方交通大学出版社.2003:158-159.
[30]王良红,冷建华.汽车倾斜牌照中字符的定位与提取[J].电讯技术,2003,4:61.
[31]刘志勇,刘迎建.车牌识别中的图像提取及分割[J].中文信息学报.2000.4
[32]王慧燕.图像边缘检测和图像匹配研究及应用[D].浙江大学博士论文2003,8
[33]刘宁.指纹匹配算法研究[D].山东大学硕士论文2006.4.
[34]龙世渚,黄载禄,曹汉房.一种立体视觉图像的特征选择与匹配方法[J].华中理工大学学报,1994,22(11):28-31
[35]舒丽霞,周成平,彭晓明,丁明跃.基于Hausdorff距离图像配准方法研究[J].中国图像图形学报.2003.12.第8卷(A版)第12期
[36]李震,范湘涛,施建成.基于分窗口相关的遥感图像配准方法[J].中国图像图形学报.2001.02,第6卷(A版)第2期
[37]李秀秀,郑江滨,张艳宁.一种新的自动图像配准技术[J].计算机应用研究.2008.1.第25卷第1期
[38]陈世伟,李世平,管京周,熊楠.一种改进的基于轮廓的图像配准算法[J].计量技术.2005.9.
[39]荣江,王文杰,陈建华.智能车牌识别系统设计与实现[J].微型机与应用.2005.01
[40]管业鹏,童林夙.二维图像特征点快速提取算法[J].中国图像图形学报.2002.12.第7卷(A版)第12期
[41]郑宇,肖南峰.一种基于相似度的汽车车牌号码自动识别系统[J].交通与计算机.2004.第5期,第22卷
[42]杨威,赵剡,许东.基于人眼视觉的结构相似度图像质量评价方法[J].北京航空航天大学学报.2008.第34卷第1期
[43]张健,毛二可,赵保军,王靖.真假目标红外图像相似度检测算法[J].光电工程.2007.第34卷第9期