基于BF531的数字仪表图像识别系统研究
摘要
随着人们生活水平的不断提高，在计量行业里的燃气表、煤气表、水表等数字仪表已经得到了越来越广泛的应用，同时这类数字仪表是否能够真实准确地反映出人们对天然气、煤气或者自然水的使用情况，关系着消费者和供应者双方面的利益。以燃气表为例，燃气表误差检定装置就是一类能够对燃气表做误差检定的设备，使用它能准确地检测出燃气表是否为合格产品。在这类数字仪表的误差检定装置中，需要对数字仪表转过的圈数做采样，从而计算数字仪表的误差，传统的做法是使用光电采样器等方式来测量数字仪表转过的圈数，然而光电采样器对一部分数字仪表是无法检测到它的读数的，不能具有通用性。
本文针对数字仪表图像的特点，创新地提出了一种基于BF531处理器的图像处理系统实现对数字仪表的识别。系统利用图像处理技术，实现对数字仪表图像的识别，来获取仪表的测量结果，实现对数字仪表的智能控制。系统首先在PC机上实现图像识别算法，然后将其移植到BF531处理器中。系统要检定一块数字仪表是否合格，可以设定仪表转过一圈表示流过的气体体积为一固定值，在检测表误差的过程中，记录当表转过一圈时表流过的体积，来计算数字仪表的误差。
本文主要工作如下：
(1)针对系统的设计要求，采用了ADI公司生产的BF531芯片作为核心处理器，该处理器的使用让项目的生产成本降低，同时对算法的处理性能各个方面也都达到了最优的效果。
(2)介绍了系统的硬件平台，给出了系统的电路原理图和PCB版图。
(3)对数字仪表图像的识别算法做了大量的实验，分别包括数字仪表图像的运动背景检测和背景建模算法、细节点和特征点匹配算法都做分析比较，发现算法的性能和复杂度方面都不适用于本系统。最后使用了基于L2范数距离最小化的图像识别算法，在完成对数字仪表图像识别的同时，表图像处理的性能也符合设计的要求，同时算法的复杂度也非常小。
(4)在VDSP++5.0开发环境下，完成了系统的软件设计，以及算法的移植和优化，算法的时间消耗达到了实时处理数字仪表图像的能力。
经过系统的整机调试和测试数据分析，基于BF531处理器的数字仪表图像识别系统能实时准确地对数字仪表转过的圈数进行采样，为数字仪表误差检定装置提供可靠的参数支持。
With the quickly development of people living standard,Gas meter, water meter or other digitalinstruments in the measurement of industry have been used more and more widely. Whether thiskind of digital instrument is capable of accurately reflects usage of the gas and water, it's related totheir benefit both the consumer and the supplier.As an example of the gas meter,gas metercalibration device is a kind of device that can examine the gas meter's error, and it can accuratelyexamine the meter whether is the certified product.In this kind of digital instrument's errorexamines,it need to make the sampling to the digital instrument turned laps, thus calculating thedigital instrument error.The traditional approach to measure the laps is to use a electro-opticalsampling instrument.But the electro-optical sampling instrument is not universal,it can't examinesome digital instrument's results.
In the view of the digital instrument image's characteristic in this paper, it is proposed a novelimage processing system based on BF531 processor to realize the recognition of suchinstrument.The system is used to the digital image processing technology to realize therecognition,and gains the measuring results of instrument to realize to digital instrument's intelligentcontrol.At first, the system realizes the image recognition algorithm on PC machine, and then thealgorithm will be transplant to the BF531 processor.Whether a digital instrument is qualified,we canset a fixed value through the gas volume when the instrument turned a circle, and record the actualvolume when the instrument turned a circle, thus we can calculate the error.
The main work of this paper are as follows:
(1) According to the design requirements of the system, the system uses BF531 chip from theADI company as the core processor.BF531 of the system can reduce the production cost, and havereached the optimal effect on the algorithms of processing performance.
(2) The paper introduces the system's hardware platform, and has given the electric circuitschematic diagram and the PCB domain.
(3) The system does a large number of experiments, analyzes and compares on imagerecognition algorithm, includes the motion detection background, background modelingalgorithm ,minutiae and feature point matching algorithm on digital instrument image.And we findthese algorithm are not suitable for this system on the performance and complexity aboutalgorithms.At last, the system uses based on L2-norm distance minimization algorithm in imagerecognition.With the completion of image recognition of the digital instrument,the image processingperformance of the system meets the design requirements, and the complexity of the algorithm is very small.
(4) In VDSP++5.0 development environment, it has completed the system's software design,aswell as the algorithm transplant and the optimization.The time consumption of algorithm hasachieved the real-time processing digital instrument image.
Through the system debugging and testing on data, the image recognition system in digitalinstrument based on BF531 can make the sampling to the digital instrument real-time and accurately,and provide reliable parameter support for the digital instrument error calibration device.
本文针对数字仪表图像的特点，创新地提出了一种基于BF531处理器的图像处理系统实现对数字仪表的识别。系统利用图像处理技术，实现对数字仪表图像的识别，来获取仪表的测量结果，实现对数字仪表的智能控制。系统首先在PC机上实现图像识别算法，然后将其移植到BF531处理器中。系统要检定一块数字仪表是否合格，可以设定仪表转过一圈表示流过的气体体积为一固定值，在检测表误差的过程中，记录当表转过一圈时表流过的体积，来计算数字仪表的误差。
本文主要工作如下：
(1)针对系统的设计要求，采用了ADI公司生产的BF531芯片作为核心处理器，该处理器的使用让项目的生产成本降低，同时对算法的处理性能各个方面也都达到了最优的效果。
(2)介绍了系统的硬件平台，给出了系统的电路原理图和PCB版图。
(3)对数字仪表图像的识别算法做了大量的实验，分别包括数字仪表图像的运动背景检测和背景建模算法、细节点和特征点匹配算法都做分析比较，发现算法的性能和复杂度方面都不适用于本系统。最后使用了基于L2范数距离最小化的图像识别算法，在完成对数字仪表图像识别的同时，表图像处理的性能也符合设计的要求，同时算法的复杂度也非常小。
(4)在VDSP++5.0开发环境下，完成了系统的软件设计，以及算法的移植和优化，算法的时间消耗达到了实时处理数字仪表图像的能力。
经过系统的整机调试和测试数据分析，基于BF531处理器的数字仪表图像识别系统能实时准确地对数字仪表转过的圈数进行采样，为数字仪表误差检定装置提供可靠的参数支持。
With the quickly development of people living standard,Gas meter, water meter or other digitalinstruments in the measurement of industry have been used more and more widely. Whether thiskind of digital instrument is capable of accurately reflects usage of the gas and water, it's related totheir benefit both the consumer and the supplier.As an example of the gas meter,gas metercalibration device is a kind of device that can examine the gas meter's error, and it can accuratelyexamine the meter whether is the certified product.In this kind of digital instrument's errorexamines,it need to make the sampling to the digital instrument turned laps, thus calculating thedigital instrument error.The traditional approach to measure the laps is to use a electro-opticalsampling instrument.But the electro-optical sampling instrument is not universal,it can't examinesome digital instrument's results.
In the view of the digital instrument image's characteristic in this paper, it is proposed a novelimage processing system based on BF531 processor to realize the recognition of suchinstrument.The system is used to the digital image processing technology to realize therecognition,and gains the measuring results of instrument to realize to digital instrument's intelligentcontrol.At first, the system realizes the image recognition algorithm on PC machine, and then thealgorithm will be transplant to the BF531 processor.Whether a digital instrument is qualified,we canset a fixed value through the gas volume when the instrument turned a circle, and record the actualvolume when the instrument turned a circle, thus we can calculate the error.
The main work of this paper are as follows:
(1) According to the design requirements of the system, the system uses BF531 chip from theADI company as the core processor.BF531 of the system can reduce the production cost, and havereached the optimal effect on the algorithms of processing performance.
(2) The paper introduces the system's hardware platform, and has given the electric circuitschematic diagram and the PCB domain.
(3) The system does a large number of experiments, analyzes and compares on imagerecognition algorithm, includes the motion detection background, background modelingalgorithm ,minutiae and feature point matching algorithm on digital instrument image.And we findthese algorithm are not suitable for this system on the performance and complexity aboutalgorithms.At last, the system uses based on L2-norm distance minimization algorithm in imagerecognition.With the completion of image recognition of the digital instrument,the image processingperformance of the system meets the design requirements, and the complexity of the algorithm is very small.
(4) In VDSP++5.0 development environment, it has completed the system's software design,aswell as the algorithm transplant and the optimization.The time consumption of algorithm hasachieved the real-time processing digital instrument image.
Through the system debugging and testing on data, the image recognition system in digitalinstrument based on BF531 can make the sampling to the digital instrument real-time and accurately,and provide reliable parameter support for the digital instrument error calibration device.
引文
[1]涂序彦,钟骥,王忠民,王沁.智能化、集成化、协调化——工业控制与自动化发展方向[J].微计算机信息,1997,13(1):4-7.
[2] Dongyuan Ge,Xifan Yao,Qing Zhang. Development of machine vision system based on BP neural network selflearning[C].Proceedings of the International Conference on Computer Science and Information Technology,2008,pages 632-636.
[3]张五,赵强松,王东云.机器视觉的现状及发展趋势[J].中原工学院学报,2008,19(1):9-1.
[4]杨铮.IC卡燃气表的现状和发展趋势[J].煤气与热力,2007,27(1):34-36.
[5]裘祖荣,石照耀,李岩.机械制造领域测量技术的发展研究[J].机械工程学报,2010,46(14):1-11.
[6]林玉池.测量控制与仪器仪表前沿技术及发展趋势[M].天津:天津大学出版社,2008.
[7] Sheeba,O.,Kumar,A. Sukesh.. Automated diagnosis of macular degeneration through image processing.Journal of the Institution of Engineers,2008,vol.89,pages 3-6.
[8]锦程.自动识别技术最新发展动态[J].水路运输文摘.2004,2(3).
[9]候海滨,沈希忠.车辆牌照识别技术的研究发展[J].上海应用技术学院学报,2009,9(3).240-243.
[10] Ratree Juntanasub,Nidapan Sureerattanan.Car license plate recognition through hausdorff distance technique[A].17thIEEE International Conference on Tool with Artificial Intelligence[C],2005,647-651.
[11]王方智.条码自动识别技术的新发展[J].物流技术与应用,2000,1:23-28.
[12] Quek.,C.,Tan,K.B.and Sagar,V.K.. Pseudo-outer product based fuzzy neural network fingerprint verification system.Neural Networks,2001,vol.4,pages 305-323.
[13]李见为.指纹识别技术的研究和发展[J].重庆职业技术学院学报,2004,13(2):4-6
[14]付同堂,苏秀琴等.基于MFC的图像处理程序的设计研究[J].科学技术与工程,2007,7(15):3690-3693,3760.
[15]孙朝云,禇燕利等.基于VC++路面裂缝图像处理系统研究[J].计算机应用软件,2009,26(8):82-85.
[16]冯冬青,付巍,石成辉.基于像素比例法和GA-BP的单相电表读书识别[J].计算机工程与设计,2009,30(4):3458-3461,3465.
[17]李健,史进.基于OpenCV的三维重建研究[J].微电子学与计算机,2008,25(12):29-32
[18]吕立新,丁德锐等.基于ARM和图像识别算法的火灾探测系统设计[J].计算机工程与设计,2008,29(10):2530-2533.
[19] Zhang Daode,Yang Guangyou,etc.Hardware design of image recognition system based ARM and FPGA[C].The 3rdIEEE conference on Industrial Electronics and Applications,2008,pages 635-638.
[20]王永欣,佟立飞,唐艺灵.基于FPGA的网络图像采集处理系统设计[J].现代电子技术,2011,34(20):29-32.
[21] Zhang Xinxi,Li Yong,etc. Design of high-speed image processing system based on FPGA[C]. The 9thInternational Conference on Electronic Measurement and Instruments,2009,pages 465-469.
[22]刘红喜,李长江,孙俊喜.基于DSP的图像采集及处理系统的设计与实现[J].电子技术应用,2009,2:22-26.
[23] Zhou Jianjun,Zhou Jianhong. Research on embedded digital image recognition system based on ARM-DSP[C].The 2nd IEEE International Conference on Computer Science and Information Technology,2009,pages 524-527.
[24]吴冬梅,张玉杰.DSP技术及应用[M].北京:北京大学出版社,2006.
[25] Analog Device. Blackfin Processor Architechture Overview[EB]. http://www.analog.com/processors/blackfin/overview/archOverview.html.
[26] Analog Device. Blackfin Embedded Processor ADSP-BF531.2006-08.
[27] Analog Device,Inc.ADSP-BF531/ADSP-BF532/ADSP-BF533 Blackfin Embedded Processor,2008.
[28]尹传历,刘冬梅,宋建中,刘太辉.基于视频图像的数字仪表读数自动识别[J].传感器与仪器仪表,2008,24(10-1):219-221.
[29]郗润平,赵荣椿,岳思聪,易飞.一个基于DSP与图像处理的水表读识系统[J].测控技术,2007,26(5):21-23.
[30]张刘胜.基于DSP的嵌入式数字仪表识别系统[D].浙江工业大学,2008.
[31]孙鲁涌,王翔,刘庆民.基于构成模式的煤气表读数字符识别技术研究[J].现代制造工程,2007,11:103-105.
[32]张艰,赵宇明,陈德权.分时电表表盘数字读数的自动识别系统[J].计算机工程,2005,31(5):178-180
[33]范新南,郭建甲,苏丽媛.基于数学形态学的数字仪表数码识别快速算法[J].计算机测量与控制,2006,14(11):1589-1590,1593.
[34]卢卫娜,刘长荣,郑玉才,王海芳.一种基于模板匹配的数字仪表字符识别方法[J].现代计算机,2008,279:70-72,86.
[35] Atsushi Shimada,Rin-ichiro Taniguchi.Hybrid Background Model using Spatial-Temporal LBP[C].2009 Advanced Video and Signal Based Surveillance,2009,pages 19-24.
[36] Mahfuzul Haque,Manzur Murshed,and Manoranjan Paul.A Hybrid Object Detection Technique from Dynamic Background Using Gaussian Mixture Models[C].Multimedia Signal Processing,2008,pages915-920.
[37]黄东军,冯佳.视频场景切分算法研究[J].企业技术开发,2009,5(28):1-3,9.
[38] Juan Zhu,YongPing Kong.A Fast Method For Building And UpDating Background Model[C].Intelligent Systems and Application,2009,pages 1-4.
[39]李小鹏,严严,章毓晋.若干背景建模方法的分析和比较[C].第十三届全国图像图形学学术会议,2006.
[40] Xudong Jiang,Wei-Yun Yau.Fingerprint Minutiae Matching Based on the Local And Global Structures[C].Pattern Recognition,2000(2),pages 1038-1041.
[41]赵娟,王典洪.指纹图像匹配的算法研究及其实现[J].计算机工程与应用,2005,13:66-69.
[42] Steven Gold,Anand Rangarajan. A graduated assignment algorithm for graph matching[C].Pattern Analysis and Machine Intelligence,1996,pages 377-388.
[43] G. Parziale,A. Niel. A fingerprint matching using minutiae triangulation. Proc. of ICBA,2004, pages 241-248.
[44] F.A. Afsar,M.Arif,M. Hussain. Fingerprint Identification and Verification System using Minutiae Matching[C].National Conference on EmERGING technologies,2004,pages 141-146.
[45]江杰,胡晓莉,李杰.一种新的指纹细节点匹配算法的研究[J].计算机应用,2005,12(5):114-115.
[46] Xuefeng Liang,Tetsuo Asano,Arijit Bishnu. Fingerprint Matching Using Minutia Polygons[C].Pattern Recognition,ICPR 2006,pages 1046-1049.
[47] A.M.Bazen,S.H.Gerez. Fingerprint matching by thin-plate spline modelling of elastic defromations. Pattern Recognition, 2003,36(8),pages 1859-1867.
[48]谭磊,张桦,薛彦斌.一种基于特征点的图像匹配算法[J].天津理工大学学报,2006,22(6):66-69.
[49] Luo Xiping,Jie Tian,Yan Wu. A minutia matching algorithm in fingerprint verification[C].Proceedings of15th ICPR,2000,pages 833-836.
[50] Demirel,H.,Anbarjafari,G..A new face recognition system based on color histogram matching[C].Communication and Applications Conference,2008,pages 1-4.
[51] Ying Fang,Huiyuan Wang,etc.Multi-object Trcaking Based on Region Corresponding and Improved Color-Histogram Matching[C].Signal Processing and Information Technology,2007,pages 1-4.
[52] Hongliang Fei and Jun Huan. L2 norm regularized feature kernel regression for graph data[C]. Conference on Information and Knowledge Management,2009,pages 593-600.
[53] Ming-Jung Seow,Valaparla,D. And Asari,V.k. L2-norm approximation based learning in recurrent neural networks for expression invariant face recognition[C].Systems,Man and Cybernetics,2003,vol.4,pages3512-3517.
[54] S.Shen,W. Shi and Y. Liu. Monocular 3D tracking of inextensible deformable surfaces under L2-norm.IEEE Trans.Image Processing,2010,pages 512-521.
[55] Djouadi,S. M. Comment on "Is the Frobenius Matrix Norm Induced?"and reply.IEEE Automatic Control,2003,vol.48,pages 518-519.
[56] Thomas Huckle,Alexander Kallischko. Frobenius Norm Minimization and Probing for Preconditioning.International Journal of Computer Mathematics,2007,vol.0,pages 1-31.
[57] Sebahattin Bektas and Yasemin Sisman. The comparison of L1 and L2-norm minimization methods.International Journal of the Physical Science,2010,vol.5(11),pages 1721-1727.
[58] Levente Kovács,Levente Kovács,etc. Induced L2-norm minimization of glucose–insulin system for Type I diabetic patients.Computer Methods and Programs in Biomedicine,2011,vol.102,pages 105-118.
[59] Bhowmik,U.K.,Ashrafi,A.and Adhami,R.R.. A Fingerprint Verification Algotithm Using the smallest Minimum Sum of Closest Euclidean Distance[C].International Conference on Electrical,Communications,and Computers,2009,pages 90-95.
[60]边肇琪,张学工.模式识别.第三版[M].清华大学出版社,2000.
[61]周海云,陈东青等.矩阵理论简明教程[M].北京:国防工业出版社,2011.
[ 62] L2-norm.URL:http://mathworld.wolfram.com/L2-Norm.html.
[63]孙燮华.数字图像处理原理与算法[M].北京:机械工业出版社,2010.
[64]于仕琪,刘瑞桢.学习OpenCV(中文版)[M].北京:清华大学出版社,2009.
[65]陈峰. Blackfin系列DSP原理与系统设计[M].北京:电子工业出版社,2004.
[66]陈峰.基于Blackfin DSP的数字图像处理[M].北京:电子工业出版社,2009.
[67]陈峰.Blackfin系列DSP原理与系统设计(第二版)[M].北京:电子工业出版社,2010.
[68]罗勇江,刘书明,肖科.Visual DSP++集成开发环境实用指南[M].北京:电子工业出版社,2008.
[2] Dongyuan Ge,Xifan Yao,Qing Zhang. Development of machine vision system based on BP neural network selflearning[C].Proceedings of the International Conference on Computer Science and Information Technology,2008,pages 632-636.
[3]张五,赵强松,王东云.机器视觉的现状及发展趋势[J].中原工学院学报,2008,19(1):9-1.
[4]杨铮.IC卡燃气表的现状和发展趋势[J].煤气与热力,2007,27(1):34-36.
[5]裘祖荣,石照耀,李岩.机械制造领域测量技术的发展研究[J].机械工程学报,2010,46(14):1-11.
[6]林玉池.测量控制与仪器仪表前沿技术及发展趋势[M].天津:天津大学出版社,2008.
[7] Sheeba,O.,Kumar,A. Sukesh.. Automated diagnosis of macular degeneration through image processing.Journal of the Institution of Engineers,2008,vol.89,pages 3-6.
[8]锦程.自动识别技术最新发展动态[J].水路运输文摘.2004,2(3).
[9]候海滨,沈希忠.车辆牌照识别技术的研究发展[J].上海应用技术学院学报,2009,9(3).240-243.
[10] Ratree Juntanasub,Nidapan Sureerattanan.Car license plate recognition through hausdorff distance technique[A].17thIEEE International Conference on Tool with Artificial Intelligence[C],2005,647-651.
[11]王方智.条码自动识别技术的新发展[J].物流技术与应用,2000,1:23-28.
[12] Quek.,C.,Tan,K.B.and Sagar,V.K.. Pseudo-outer product based fuzzy neural network fingerprint verification system.Neural Networks,2001,vol.4,pages 305-323.
[13]李见为.指纹识别技术的研究和发展[J].重庆职业技术学院学报,2004,13(2):4-6
[14]付同堂,苏秀琴等.基于MFC的图像处理程序的设计研究[J].科学技术与工程,2007,7(15):3690-3693,3760.
[15]孙朝云,禇燕利等.基于VC++路面裂缝图像处理系统研究[J].计算机应用软件,2009,26(8):82-85.
[16]冯冬青,付巍,石成辉.基于像素比例法和GA-BP的单相电表读书识别[J].计算机工程与设计,2009,30(4):3458-3461,3465.
[17]李健,史进.基于OpenCV的三维重建研究[J].微电子学与计算机,2008,25(12):29-32
[18]吕立新,丁德锐等.基于ARM和图像识别算法的火灾探测系统设计[J].计算机工程与设计,2008,29(10):2530-2533.
[19] Zhang Daode,Yang Guangyou,etc.Hardware design of image recognition system based ARM and FPGA[C].The 3rdIEEE conference on Industrial Electronics and Applications,2008,pages 635-638.
[20]王永欣,佟立飞,唐艺灵.基于FPGA的网络图像采集处理系统设计[J].现代电子技术,2011,34(20):29-32.
[21] Zhang Xinxi,Li Yong,etc. Design of high-speed image processing system based on FPGA[C]. The 9thInternational Conference on Electronic Measurement and Instruments,2009,pages 465-469.
[22]刘红喜,李长江,孙俊喜.基于DSP的图像采集及处理系统的设计与实现[J].电子技术应用,2009,2:22-26.
[23] Zhou Jianjun,Zhou Jianhong. Research on embedded digital image recognition system based on ARM-DSP[C].The 2nd IEEE International Conference on Computer Science and Information Technology,2009,pages 524-527.
[24]吴冬梅,张玉杰.DSP技术及应用[M].北京:北京大学出版社,2006.
[25] Analog Device. Blackfin Processor Architechture Overview[EB]. http://www.analog.com/processors/blackfin/overview/archOverview.html.
[26] Analog Device. Blackfin Embedded Processor ADSP-BF531.2006-08.
[27] Analog Device,Inc.ADSP-BF531/ADSP-BF532/ADSP-BF533 Blackfin Embedded Processor,2008.
[28]尹传历,刘冬梅,宋建中,刘太辉.基于视频图像的数字仪表读数自动识别[J].传感器与仪器仪表,2008,24(10-1):219-221.
[29]郗润平,赵荣椿,岳思聪,易飞.一个基于DSP与图像处理的水表读识系统[J].测控技术,2007,26(5):21-23.
[30]张刘胜.基于DSP的嵌入式数字仪表识别系统[D].浙江工业大学,2008.
[31]孙鲁涌,王翔,刘庆民.基于构成模式的煤气表读数字符识别技术研究[J].现代制造工程,2007,11:103-105.
[32]张艰,赵宇明,陈德权.分时电表表盘数字读数的自动识别系统[J].计算机工程,2005,31(5):178-180
[33]范新南,郭建甲,苏丽媛.基于数学形态学的数字仪表数码识别快速算法[J].计算机测量与控制,2006,14(11):1589-1590,1593.
[34]卢卫娜,刘长荣,郑玉才,王海芳.一种基于模板匹配的数字仪表字符识别方法[J].现代计算机,2008,279:70-72,86.
[35] Atsushi Shimada,Rin-ichiro Taniguchi.Hybrid Background Model using Spatial-Temporal LBP[C].2009 Advanced Video and Signal Based Surveillance,2009,pages 19-24.
[36] Mahfuzul Haque,Manzur Murshed,and Manoranjan Paul.A Hybrid Object Detection Technique from Dynamic Background Using Gaussian Mixture Models[C].Multimedia Signal Processing,2008,pages915-920.
[37]黄东军,冯佳.视频场景切分算法研究[J].企业技术开发,2009,5(28):1-3,9.
[38] Juan Zhu,YongPing Kong.A Fast Method For Building And UpDating Background Model[C].Intelligent Systems and Application,2009,pages 1-4.
[39]李小鹏,严严,章毓晋.若干背景建模方法的分析和比较[C].第十三届全国图像图形学学术会议,2006.
[40] Xudong Jiang,Wei-Yun Yau.Fingerprint Minutiae Matching Based on the Local And Global Structures[C].Pattern Recognition,2000(2),pages 1038-1041.
[41]赵娟,王典洪.指纹图像匹配的算法研究及其实现[J].计算机工程与应用,2005,13:66-69.
[42] Steven Gold,Anand Rangarajan. A graduated assignment algorithm for graph matching[C].Pattern Analysis and Machine Intelligence,1996,pages 377-388.
[43] G. Parziale,A. Niel. A fingerprint matching using minutiae triangulation. Proc. of ICBA,2004, pages 241-248.
[44] F.A. Afsar,M.Arif,M. Hussain. Fingerprint Identification and Verification System using Minutiae Matching[C].National Conference on EmERGING technologies,2004,pages 141-146.
[45]江杰,胡晓莉,李杰.一种新的指纹细节点匹配算法的研究[J].计算机应用,2005,12(5):114-115.
[46] Xuefeng Liang,Tetsuo Asano,Arijit Bishnu. Fingerprint Matching Using Minutia Polygons[C].Pattern Recognition,ICPR 2006,pages 1046-1049.
[47] A.M.Bazen,S.H.Gerez. Fingerprint matching by thin-plate spline modelling of elastic defromations. Pattern Recognition, 2003,36(8),pages 1859-1867.
[48]谭磊,张桦,薛彦斌.一种基于特征点的图像匹配算法[J].天津理工大学学报,2006,22(6):66-69.
[49] Luo Xiping,Jie Tian,Yan Wu. A minutia matching algorithm in fingerprint verification[C].Proceedings of15th ICPR,2000,pages 833-836.
[50] Demirel,H.,Anbarjafari,G..A new face recognition system based on color histogram matching[C].Communication and Applications Conference,2008,pages 1-4.
[51] Ying Fang,Huiyuan Wang,etc.Multi-object Trcaking Based on Region Corresponding and Improved Color-Histogram Matching[C].Signal Processing and Information Technology,2007,pages 1-4.
[52] Hongliang Fei and Jun Huan. L2 norm regularized feature kernel regression for graph data[C]. Conference on Information and Knowledge Management,2009,pages 593-600.
[53] Ming-Jung Seow,Valaparla,D. And Asari,V.k. L2-norm approximation based learning in recurrent neural networks for expression invariant face recognition[C].Systems,Man and Cybernetics,2003,vol.4,pages3512-3517.
[54] S.Shen,W. Shi and Y. Liu. Monocular 3D tracking of inextensible deformable surfaces under L2-norm.IEEE Trans.Image Processing,2010,pages 512-521.
[55] Djouadi,S. M. Comment on "Is the Frobenius Matrix Norm Induced?"and reply.IEEE Automatic Control,2003,vol.48,pages 518-519.
[56] Thomas Huckle,Alexander Kallischko. Frobenius Norm Minimization and Probing for Preconditioning.International Journal of Computer Mathematics,2007,vol.0,pages 1-31.
[57] Sebahattin Bektas and Yasemin Sisman. The comparison of L1 and L2-norm minimization methods.International Journal of the Physical Science,2010,vol.5(11),pages 1721-1727.
[58] Levente Kovács,Levente Kovács,etc. Induced L2-norm minimization of glucose–insulin system for Type I diabetic patients.Computer Methods and Programs in Biomedicine,2011,vol.102,pages 105-118.
[59] Bhowmik,U.K.,Ashrafi,A.and Adhami,R.R.. A Fingerprint Verification Algotithm Using the smallest Minimum Sum of Closest Euclidean Distance[C].International Conference on Electrical,Communications,and Computers,2009,pages 90-95.
[60]边肇琪,张学工.模式识别.第三版[M].清华大学出版社,2000.
[61]周海云,陈东青等.矩阵理论简明教程[M].北京:国防工业出版社,2011.
[ 62] L2-norm.URL:http://mathworld.wolfram.com/L2-Norm.html.
[63]孙燮华.数字图像处理原理与算法[M].北京:机械工业出版社,2010.
[64]于仕琪,刘瑞桢.学习OpenCV(中文版)[M].北京:清华大学出版社,2009.
[65]陈峰. Blackfin系列DSP原理与系统设计[M].北京:电子工业出版社,2004.
[66]陈峰.基于Blackfin DSP的数字图像处理[M].北京:电子工业出版社,2009.
[67]陈峰.Blackfin系列DSP原理与系统设计(第二版)[M].北京:电子工业出版社,2010.
[68]罗勇江,刘书明,肖科.Visual DSP++集成开发环境实用指南[M].北京:电子工业出版社,2008.