基于DSP的车牌号识别技术及应用研究
|
摘要
随着我国交通事业的迅速发展,越来越多的人拥有了汽车,国内高速公路、城市道路、停车场的建设也日益增多,这些情况催生了智能交通系统ITS(IntelligentTraffic System)的出现。车牌牌照识别(License Plate Recognition,LPR)系统是智能交通系统的一个重要子系统,其本身也是一个整体的智能化解决方案,主要包括图像采集、车牌定位、字符分割、字符识别等模块。近年来,车牌识别系统开始应用于小区的车辆管理,公安系统的车辆跟踪、案件侦破等方面。相比传统的PC机纯软件车牌识别方式,DSP嵌入式车牌识别系统在体积、重量、功耗以及图像处理速度等方面都具有明显的优势,成为众多用户的首选方案。本文从车牌识别算法的研究和硬件平台的构建两个方面对DSP车牌识别技术做了比较深入的研究。
论文首先阐述了图像的一些基本知识,其中比较重要的知识点是颜色空间和图像数字化及其存储技术。然后详细论述了车牌识别中的每个步骤的具体实现方法,提出了充分利用车牌及车牌字符的先验知识定位车牌和分割车牌图像的方法,在之后的车牌字符识别中提出了二重模板匹配法,有效提高了车牌字符识别的正确率。接着论述了系统硬件平台的设计,包括视频采集、电源、复位、RAM/FLASH扩展、JTAG仿真和调试等电路模块。最后,介绍了DSP的有关开发技术,包括CCS集成开发环境的使用、DSP/BIOS的配置和使用,代码的优化等。
实验表明,本文提出的车牌号识别方法具有比较高的正确识别率和较快的识别速度,具有较高的实用价值。
文章最后,总结了本文所做的工作,并给出了未来的研究方向。
With the rapid development of our country's traffic, more and more persons possess of cars. The speedways, roadways and stopping places are increasing everyday. All the phenomena result in the appearing of Intelligent Traffic System. License Plate Recognition is an important subsystem of the Intelligent Traffic System. It is also an integrated intelligent project. It includes image collection, license plate location, image segmentation, character recognizing. These years, LPR system begins to be used in vehicle management in uptown, vehicle track and base reconnaissance in police department. Compared with PC LPR system, the embedded DSP LPR system has more advantages in size, weight, power consumption and image processing rate. So, it has become the first choose of more users.
First of all, this paper introduces some basic knowledge of image. The two important facets is color space, image digit transformation and storage technology. Afterwards, this paper dissertates the realization ways of every process of LPR technology and proposes the ways of license plate location and license plate image segmentation which adequately use license plate's characteristics to realize. It also proposes the way of template matched twice which used to divide up the character images, it enhanced the character recognizing effect. In succession, this paper dissertates the designs of every hardware module of this DSP LPR system witch include video collection, power, reset, RAM/FLASH expending, JTAG and circuit debugging module. Finally, the paper dissertates DSP exploitation technology which including the CCS integrated exploitation environment, the configuring and using of DSP/BIOS, and the optimization of code.
The experimentations indicate that the ways of LPR proposed in this paper achieve a good recognizing effect and recognizing speed.
In the end, it summarizes the tasks which accomplish by this paper and proposes some works which should be consummated by the research in the future.
论文首先阐述了图像的一些基本知识,其中比较重要的知识点是颜色空间和图像数字化及其存储技术。然后详细论述了车牌识别中的每个步骤的具体实现方法,提出了充分利用车牌及车牌字符的先验知识定位车牌和分割车牌图像的方法,在之后的车牌字符识别中提出了二重模板匹配法,有效提高了车牌字符识别的正确率。接着论述了系统硬件平台的设计,包括视频采集、电源、复位、RAM/FLASH扩展、JTAG仿真和调试等电路模块。最后,介绍了DSP的有关开发技术,包括CCS集成开发环境的使用、DSP/BIOS的配置和使用,代码的优化等。
实验表明,本文提出的车牌号识别方法具有比较高的正确识别率和较快的识别速度,具有较高的实用价值。
文章最后,总结了本文所做的工作,并给出了未来的研究方向。
With the rapid development of our country's traffic, more and more persons possess of cars. The speedways, roadways and stopping places are increasing everyday. All the phenomena result in the appearing of Intelligent Traffic System. License Plate Recognition is an important subsystem of the Intelligent Traffic System. It is also an integrated intelligent project. It includes image collection, license plate location, image segmentation, character recognizing. These years, LPR system begins to be used in vehicle management in uptown, vehicle track and base reconnaissance in police department. Compared with PC LPR system, the embedded DSP LPR system has more advantages in size, weight, power consumption and image processing rate. So, it has become the first choose of more users.
First of all, this paper introduces some basic knowledge of image. The two important facets is color space, image digit transformation and storage technology. Afterwards, this paper dissertates the realization ways of every process of LPR technology and proposes the ways of license plate location and license plate image segmentation which adequately use license plate's characteristics to realize. It also proposes the way of template matched twice which used to divide up the character images, it enhanced the character recognizing effect. In succession, this paper dissertates the designs of every hardware module of this DSP LPR system witch include video collection, power, reset, RAM/FLASH expending, JTAG and circuit debugging module. Finally, the paper dissertates DSP exploitation technology which including the CCS integrated exploitation environment, the configuring and using of DSP/BIOS, and the optimization of code.
The experimentations indicate that the ways of LPR proposed in this paper achieve a good recognizing effect and recognizing speed.
In the end, it summarizes the tasks which accomplish by this paper and proposes some works which should be consummated by the research in the future.
引文
[1]陆化普,史其信.中国ITS发展战略构思.公路交通科技,1998,15(3):13-16
[2]戴东昌,蔡建华.国外解决城市交通拥堵问题对策.求是,2004,23
[3]Shyang-Lih Chang;Li-Shien Chen;Yun-Chung Chung;Sei-Wan Chen,Automatic license plate recognition.Intelligent Transportation Systems,IEEE Transactions on Volume 5,Issue 1,March 2004 Page(s):42-53.
[4]孙启星.车牌识别系统的研究与实现:[硕士学位论文].哈尔滨:哈尔滨工程大学,2005:2-4
[5]Rafael C.Gonzalez,Richard E.Woods,Digital Image Processing(Prentice-Hall Inc,2002).
[6]Aghdasi,F.;Ndungo,H.,Automatic licence plate recognition system,AFRICON,2004.7th AFRICON Conference in Africa Volume 1,15-17 Sept.2004 Page(s):45-50Vol.1.
[7]Taleb-Ahmed,A.Hamad,D.Tilmant.Vehicle license plate recognition in marketing application.Intelligent Vehicles Symposium.IEEE,9-11 June 2003 Pages:90-94
[8]Hi-Tech Solution有限公司网站.http://www.htsol.com
[9]Optasia Systems Pte有限公司网站.http://www.singaporegateway.com/optasia/
[10]傅得胜.图形图像处理学.南京:东南大学出版社,2001
[11]陈书海,傅录祥.实用数字图像处理.北京:科学出版社,2000
[12]章毓晋.图像工程.北京:清华大学出版社,1999
[13]孟章荣.各种颜色模型选用需求分析.中国图像图形学报,1996.1(3):238-241
[13]何铁军,张宁,黄卫.车牌识别算法的研究与实现.公路交通科技,2006,(8):147-149
[14]一种新的车牌识别预处理算法.西北工业大学,2002.20(1):83-86
[15]贾永红.数字图像处理.武汉:武汉大学出版社,2003
[16]贾云得.机器视觉.北京:科学出版社,2000
[17]吴明赞,王志功.基于ADSP的图像增强算法实现.计算机应用研究.2006
[18]赵俊霞.对图像进行直方图均衡处理的DSP芯片设计.中国医学物理学杂志,第23卷第2期
[19]吴传孙,邹扬德,周定康.二值化算法在车牌识别中的应用.计算机与现代化,2003(6):13-16
[20]孙启星.车牌识别系统的研究与实现.哈尔滨工程大学,2005:2-4
[21]基于颜色和纹理分析的车牌定位新方法.中国图像图形学报,2001.10
[22]赵海涛,周建波,董介春.基于TMS320C6416的Sobel算法图像边缘检测.青岛大学学报(工程技术版),2006(2)
[23]崔屹.图像处理与分析-数学形态学方法及应用.第1版.北京:科学出版社,2000
[24]陈寅鹏,丁晓青.复杂车牌图像中的车牌定位与字符分割方法.红外与激光工程,2004.01
[25]迟晓君,孟庆春.基于投影特征值的车牌字符分割算法.计算机应用研究,2006,23(7):256-257
[26]杨晓梅,吴炜,黎涛,何小海.基于Gabor变换和支持向量机的车牌识别算法.四川大学学报工程科学版,2005,37(5):130-135
[27]马俊莉,莫玉龙,王明祥.一种基于改进模板匹配的车牌字符识别方法.小型微型计算机系统,2003.9
[28]魏武,黄心汉等.基于模板匹配和神经网络的车牌网络的车牌字符识别方法.模式识别与人工智能,2001.14(1):123-126
[29]潘翔,叶修梓,张三元.车牌汉字特征提取.中国图像图形学报,2003,8(10):1218-1222
[30]周霖.DSP通信工程技术应用.北京:国防工业出版社,2004
[31]李方慧,王飞,何佩琨.TMS320C6000系列DSPs原理与应用.电子工业出版社,2002
[32]Texas Instrument.TMS320DM642 Video Port Mini-Driver.2003
[33]Texas Instrament.TMS320DM642 Video/Imaging Fixed-Point Digital Signal Proeessor.2005
[34]Philips Semiconductors.DatasheetSAA7113H CVIP2.2001
[35]Philips Semiconductors,SAA7113H 9-bit video input processor,1999
[36]SAA7111A Enhanced Video Input Processor Datasheet.1998
[37]Texas Instruments.TMS320C6000 DSP Inter-lntegrated Circuit(12C)Module Reference Guide,2002
[38]廖良斌,喻方平.基于DSP和USB的图像采集系统的研究.武汉理工大学学报(交通科学与工程版),2006(1)
[39]刘源,朱善安,叶旭东.基于DM642的嵌入式视频监控系统硬件设计.电子元器件应用,2006,29(3):905-908
[40]王应军,杨国胜,范秋风,马晓燕.DSP+FPGA的实时图像处理硬件系统设计,河南大学学报,2006(1)
[41]Texas Instrument.TMS320C6000 EMIF to External SDRAM Interface.2001
[42]AMD.Am29LV640MT/B DataSheet.2003-7
[43]Texas Instruments.TMS320C6000 EMIF-to-External Flash Memory,2002 Philips Semiconductors.THE I2 C-BUS SPECIFICATION VERSION 2.1,2000
[44]Texas Instrument.Code Composer Studio User's Guide.SPRU328B.2002
[45]Texas Instrument.TMS320 DSP/BIOS Application Programming Interface(API)Reference Guide.2003
[46]Texas Instrument.TMS320CG000 DSP/BIOS User's Guide.2003
[47].TMS320CG000 Optimizing C/C++Compiler User's Guide,Texas Instruments Incorporated,April 2001
[48]Texas Instrument.TMS320C6000 Chip Support Library API User's Guide.2003
[49]Texas Instrument.TMS320C6000 Programmer's Guide.SPRU198F,2001
[2]戴东昌,蔡建华.国外解决城市交通拥堵问题对策.求是,2004,23
[3]Shyang-Lih Chang;Li-Shien Chen;Yun-Chung Chung;Sei-Wan Chen,Automatic license plate recognition.Intelligent Transportation Systems,IEEE Transactions on Volume 5,Issue 1,March 2004 Page(s):42-53.
[4]孙启星.车牌识别系统的研究与实现:[硕士学位论文].哈尔滨:哈尔滨工程大学,2005:2-4
[5]Rafael C.Gonzalez,Richard E.Woods,Digital Image Processing(Prentice-Hall Inc,2002).
[6]Aghdasi,F.;Ndungo,H.,Automatic licence plate recognition system,AFRICON,2004.7th AFRICON Conference in Africa Volume 1,15-17 Sept.2004 Page(s):45-50Vol.1.
[7]Taleb-Ahmed,A.Hamad,D.Tilmant.Vehicle license plate recognition in marketing application.Intelligent Vehicles Symposium.IEEE,9-11 June 2003 Pages:90-94
[8]Hi-Tech Solution有限公司网站.http://www.htsol.com
[9]Optasia Systems Pte有限公司网站.http://www.singaporegateway.com/optasia/
[10]傅得胜.图形图像处理学.南京:东南大学出版社,2001
[11]陈书海,傅录祥.实用数字图像处理.北京:科学出版社,2000
[12]章毓晋.图像工程.北京:清华大学出版社,1999
[13]孟章荣.各种颜色模型选用需求分析.中国图像图形学报,1996.1(3):238-241
[13]何铁军,张宁,黄卫.车牌识别算法的研究与实现.公路交通科技,2006,(8):147-149
[14]一种新的车牌识别预处理算法.西北工业大学,2002.20(1):83-86
[15]贾永红.数字图像处理.武汉:武汉大学出版社,2003
[16]贾云得.机器视觉.北京:科学出版社,2000
[17]吴明赞,王志功.基于ADSP的图像增强算法实现.计算机应用研究.2006
[18]赵俊霞.对图像进行直方图均衡处理的DSP芯片设计.中国医学物理学杂志,第23卷第2期
[19]吴传孙,邹扬德,周定康.二值化算法在车牌识别中的应用.计算机与现代化,2003(6):13-16
[20]孙启星.车牌识别系统的研究与实现.哈尔滨工程大学,2005:2-4
[21]基于颜色和纹理分析的车牌定位新方法.中国图像图形学报,2001.10
[22]赵海涛,周建波,董介春.基于TMS320C6416的Sobel算法图像边缘检测.青岛大学学报(工程技术版),2006(2)
[23]崔屹.图像处理与分析-数学形态学方法及应用.第1版.北京:科学出版社,2000
[24]陈寅鹏,丁晓青.复杂车牌图像中的车牌定位与字符分割方法.红外与激光工程,2004.01
[25]迟晓君,孟庆春.基于投影特征值的车牌字符分割算法.计算机应用研究,2006,23(7):256-257
[26]杨晓梅,吴炜,黎涛,何小海.基于Gabor变换和支持向量机的车牌识别算法.四川大学学报工程科学版,2005,37(5):130-135
[27]马俊莉,莫玉龙,王明祥.一种基于改进模板匹配的车牌字符识别方法.小型微型计算机系统,2003.9
[28]魏武,黄心汉等.基于模板匹配和神经网络的车牌网络的车牌字符识别方法.模式识别与人工智能,2001.14(1):123-126
[29]潘翔,叶修梓,张三元.车牌汉字特征提取.中国图像图形学报,2003,8(10):1218-1222
[30]周霖.DSP通信工程技术应用.北京:国防工业出版社,2004
[31]李方慧,王飞,何佩琨.TMS320C6000系列DSPs原理与应用.电子工业出版社,2002
[32]Texas Instrument.TMS320DM642 Video Port Mini-Driver.2003
[33]Texas Instrament.TMS320DM642 Video/Imaging Fixed-Point Digital Signal Proeessor.2005
[34]Philips Semiconductors.DatasheetSAA7113H CVIP2.2001
[35]Philips Semiconductors,SAA7113H 9-bit video input processor,1999
[36]SAA7111A Enhanced Video Input Processor Datasheet.1998
[37]Texas Instruments.TMS320C6000 DSP Inter-lntegrated Circuit(12C)Module Reference Guide,2002
[38]廖良斌,喻方平.基于DSP和USB的图像采集系统的研究.武汉理工大学学报(交通科学与工程版),2006(1)
[39]刘源,朱善安,叶旭东.基于DM642的嵌入式视频监控系统硬件设计.电子元器件应用,2006,29(3):905-908
[40]王应军,杨国胜,范秋风,马晓燕.DSP+FPGA的实时图像处理硬件系统设计,河南大学学报,2006(1)
[41]Texas Instrument.TMS320C6000 EMIF to External SDRAM Interface.2001
[42]AMD.Am29LV640MT/B DataSheet.2003-7
[43]Texas Instruments.TMS320C6000 EMIF-to-External Flash Memory,2002 Philips Semiconductors.THE I2 C-BUS SPECIFICATION VERSION 2.1,2000
[44]Texas Instrument.Code Composer Studio User's Guide.SPRU328B.2002
[45]Texas Instrument.TMS320 DSP/BIOS Application Programming Interface(API)Reference Guide.2003
[46]Texas Instrument.TMS320CG000 DSP/BIOS User's Guide.2003
[47].TMS320CG000 Optimizing C/C++Compiler User's Guide,Texas Instruments Incorporated,April 2001
[48]Texas Instrument.TMS320C6000 Chip Support Library API User's Guide.2003
[49]Texas Instrument.TMS320C6000 Programmer's Guide.SPRU198F,2001