基于显微图像的结核杆菌自动检测关键技术研究
|
摘要
结核病是由结核杆菌感染引起的慢性传染疾病，传统诊断方法存在漏检率高、劳动强度大、严重依赖专家的知识和经验等问题。为了克服诸多缺陷，本文将图像处理及模式识别技术引入结核病的诊断并实现全自动化。论文围绕结核杆菌自动检验中的显微镜自动聚焦及目标检测两大关键技术展开研究，取得了若干重要成果。
显微镜自动聚焦方面：（1）针对聚焦函数无定量评估指标的问题，设计了陡峭区宽度、平缓区波动量等6个定量评价指标，为聚焦函数的设计及选取提供了定量标准；（2）为了解决稀疏图像内容情况下传统聚焦方法聚焦成功率低的问题，提出一种基于图像内容重要度的聚焦算法。实验结果表明：当图像内容非常稀疏时，本算法聚焦成功率高达90%，而此时传统算法成功率仅为24%；（3）提出一种基于视频连续反馈的显微镜自动聚焦算法。算法首次建立了显微镜连续聚焦模型以及相关参量之间的函数约束关系，并给出了一种多视野扫描中聚焦初始方向的启发式选取方法。实验结果表明：本算法聚焦速度约为传统算法的2.2倍，聚焦成功率高达96.7%，表明本算法性能远优于传统算法。
结核杆菌识别方面：（1）建立了结核杆菌目标的颜色及形状模型，并提出一种基于目标骨架的关键点自动标注及单分支结构目标的形状简化建模方法。该方法的提出解决了经典变形模型无法自动标注关键点、无法自动获取目标形状参数的问题，从而使得大样本训练及基于变形模型的实时目标识别成为可能；（2）建立了面向实际工程应用的由粗到精、逐步细化、层层筛选的多阶段、多层次图像分割框架，并提出一种基于形状几何统计特征及决策树的结核杆菌目标识别算法。
为了验证算法性能，分别进行了典型样本实验和大样本统计实验。典型样本实验选取7组分割及识别难度比较大的典型结核杆菌显微图像，结果表明本文算法能很好地适应结核杆菌图像的背景及杂质的复杂变化，分割及识别的效果比较好。大样本实验中，从大量痰涂片样本中随机选取22836幅图像进行实验，结果为：算法灵敏度（也即真阳性率）为95.2%，特异度（也即真阴性率）为91.1%，表明本算法性能较好，但对一些极端情况还需要进一步研究。
Tuberculosis (TB) is a serious communicable chronic disease which is casued bythe Mycobacterium Tuberculosis. The conventional method employed for TB diagnosisinvolves a labor-intensive task with poor sensitivity and requires highly trained experts.To overcome these shortcomings, this dissertation employed image process and patternrecognition techniques for automatic TB diagnosis. Two key technologies employed forautomatic TB indentification, including the microscopy autofocusing technique and theobject detection technique, were studied. The technical contribution of this dissertationis summarized as follows:
First, in microscopy autofocusing the following contribution is made:(1) Wedeveloped six quantative performance metrics of focus function, including the width ofsloped part of focus curve, the steepness, the variance of flat part of the focus curve,et.al. These metrics provide the quantative standards for focus function design andperformance evaluation;(2) To overcome the shortcoming that the traditionalautofocusing algorithms may fail to find the optimal focal plane under thecircumstances with low image content densities, this dissertation proposed a contentbased focus measure for guiding automatic search of the optimal focal plane. Theexperimental results show that performance of the proposed method is far superior tothe traditional methods: the autofocusing success rate of the proposed method is largerthan90%under the circumstances with low image content density while the traditionalmethod only gains a success rate of24%;(3) A continuous autofocusing method, inwhich the z-stage keeps moving during the search process and the image exposure andthe subsequent focus measure calculation are carried out in parallel with the z-stagemovement, was proposed. This dissertation established a continuous autofocusingmodel and the relationship of the motor speed, the frame rate, and the depth of field ofthe optical system, which is essentially important for achieving continuous autofocusing.Moreover, a heuristic initial search direction selection method was proposed, whichdramatically improve the speed of autofocusing in muli-fileds scanning system.Experimental investigation has been conducted based on our automatic microscopysystem. The experimental results show that the proposed method gains a success rate of96.7%and the autofocusing speed is2.2times faster than the tranditional “stop-and-go”method. These results confirm the superior performance of the proposed method overthe traditional methods.
In automatic detection of Mycobacterium Tuberculosis (TB) in microscopicimages, we have made the following contribution:(1) A color model and a deformableshape model of the TB object were established. Moreover, a landmark automaticannotation method which is based on the morphologyical skelecton was proposed and a skelecton based simplified deformable shape model was established. The skelectonbased shape model outperforms the tranditional coutour based shape model not only forits low complexity but also for its capability that makes the landmark be auto-annotated,which makes the shape parameter auto-extraction and recognition possible;(2) A coarseto fine, multi-stage, multi-level image segmentation framework was established to dealwith the complicated situation in practical image segmentaion task. Moreover, aclassification and recognition algorithm based on the shape feature descriptors and thedecision tree classifier was proposed.
In order to demonstrate the performance of the proposed algorithms, two types ofexperiments had been condoncted, including the typical samples experiment using7typical TB images and the statistical experiments using22836TB images randomlyselected from larger number of TB specimens. Experimental results show that theproposed algorithm can accommodate the complex variety of specimens and the imagebackground. The statistical experiments result show that the sensitivity (true posistiverate) and the specificity (true negative rate) are95.2%and91%respectively. This is agood score considering that the samples used in the experiments are randomly selectedfrom TB specimens whose image quality may be influenced by many factors. In futurework, we will do futher effort to improve the performance especially the specificity ofthe algorithm.
显微镜自动聚焦方面：（1）针对聚焦函数无定量评估指标的问题，设计了陡峭区宽度、平缓区波动量等6个定量评价指标，为聚焦函数的设计及选取提供了定量标准；（2）为了解决稀疏图像内容情况下传统聚焦方法聚焦成功率低的问题，提出一种基于图像内容重要度的聚焦算法。实验结果表明：当图像内容非常稀疏时，本算法聚焦成功率高达90%，而此时传统算法成功率仅为24%；（3）提出一种基于视频连续反馈的显微镜自动聚焦算法。算法首次建立了显微镜连续聚焦模型以及相关参量之间的函数约束关系，并给出了一种多视野扫描中聚焦初始方向的启发式选取方法。实验结果表明：本算法聚焦速度约为传统算法的2.2倍，聚焦成功率高达96.7%，表明本算法性能远优于传统算法。
结核杆菌识别方面：（1）建立了结核杆菌目标的颜色及形状模型，并提出一种基于目标骨架的关键点自动标注及单分支结构目标的形状简化建模方法。该方法的提出解决了经典变形模型无法自动标注关键点、无法自动获取目标形状参数的问题，从而使得大样本训练及基于变形模型的实时目标识别成为可能；（2）建立了面向实际工程应用的由粗到精、逐步细化、层层筛选的多阶段、多层次图像分割框架，并提出一种基于形状几何统计特征及决策树的结核杆菌目标识别算法。
为了验证算法性能，分别进行了典型样本实验和大样本统计实验。典型样本实验选取7组分割及识别难度比较大的典型结核杆菌显微图像，结果表明本文算法能很好地适应结核杆菌图像的背景及杂质的复杂变化，分割及识别的效果比较好。大样本实验中，从大量痰涂片样本中随机选取22836幅图像进行实验，结果为：算法灵敏度（也即真阳性率）为95.2%，特异度（也即真阴性率）为91.1%，表明本算法性能较好，但对一些极端情况还需要进一步研究。
Tuberculosis (TB) is a serious communicable chronic disease which is casued bythe Mycobacterium Tuberculosis. The conventional method employed for TB diagnosisinvolves a labor-intensive task with poor sensitivity and requires highly trained experts.To overcome these shortcomings, this dissertation employed image process and patternrecognition techniques for automatic TB diagnosis. Two key technologies employed forautomatic TB indentification, including the microscopy autofocusing technique and theobject detection technique, were studied. The technical contribution of this dissertationis summarized as follows:
First, in microscopy autofocusing the following contribution is made:(1) Wedeveloped six quantative performance metrics of focus function, including the width ofsloped part of focus curve, the steepness, the variance of flat part of the focus curve,et.al. These metrics provide the quantative standards for focus function design andperformance evaluation;(2) To overcome the shortcoming that the traditionalautofocusing algorithms may fail to find the optimal focal plane under thecircumstances with low image content densities, this dissertation proposed a contentbased focus measure for guiding automatic search of the optimal focal plane. Theexperimental results show that performance of the proposed method is far superior tothe traditional methods: the autofocusing success rate of the proposed method is largerthan90%under the circumstances with low image content density while the traditionalmethod only gains a success rate of24%;(3) A continuous autofocusing method, inwhich the z-stage keeps moving during the search process and the image exposure andthe subsequent focus measure calculation are carried out in parallel with the z-stagemovement, was proposed. This dissertation established a continuous autofocusingmodel and the relationship of the motor speed, the frame rate, and the depth of field ofthe optical system, which is essentially important for achieving continuous autofocusing.Moreover, a heuristic initial search direction selection method was proposed, whichdramatically improve the speed of autofocusing in muli-fileds scanning system.Experimental investigation has been conducted based on our automatic microscopysystem. The experimental results show that the proposed method gains a success rate of96.7%and the autofocusing speed is2.2times faster than the tranditional “stop-and-go”method. These results confirm the superior performance of the proposed method overthe traditional methods.
In automatic detection of Mycobacterium Tuberculosis (TB) in microscopicimages, we have made the following contribution:(1) A color model and a deformableshape model of the TB object were established. Moreover, a landmark automaticannotation method which is based on the morphologyical skelecton was proposed and a skelecton based simplified deformable shape model was established. The skelectonbased shape model outperforms the tranditional coutour based shape model not only forits low complexity but also for its capability that makes the landmark be auto-annotated,which makes the shape parameter auto-extraction and recognition possible;(2) A coarseto fine, multi-stage, multi-level image segmentation framework was established to dealwith the complicated situation in practical image segmentaion task. Moreover, aclassification and recognition algorithm based on the shape feature descriptors and thedecision tree classifier was proposed.
In order to demonstrate the performance of the proposed algorithms, two types ofexperiments had been condoncted, including the typical samples experiment using7typical TB images and the statistical experiments using22836TB images randomlyselected from larger number of TB specimens. Experimental results show that theproposed algorithm can accommodate the complex variety of specimens and the imagebackground. The statistical experiments result show that the sensitivity (true posistiverate) and the specificity (true negative rate) are95.2%and91%respectively. This is agood score considering that the samples used in the experiments are randomly selectedfrom TB specimens whose image quality may be influenced by many factors. In futurework, we will do futher effort to improve the performance especially the specificity ofthe algorithm.
引文
[1]付晓燕,王学军,刘石泉等.结核分支杆菌诊断技术研究进展[J],生物技术通讯,2009,20(2):249~252.
[2] World Health Organization.Global tuberculosis control-epidemiology, strategy,financing [M]. Geneva, Switzerland, WHO Press,2009.
[3]中华人民共和国卫生部.2011年度全国法定传染病疫情概况[EB/OL].www.moh.gov.cn,2012-02-10.
[4]中国疾病预防控制中心.中国结核病防治规划实施工作指南(2008年版)[S],2008.
[5]王会莲.结核细胞自动识别系统中关键技术的研究[D].北京:北京化工大学,2006.
[6] Bello AK, Njoku CH. Tuberculosis: current trends in diagnosis and treatment [J].Niger J Clin Pract.2005,8:118~124.
[7] Negi SS, Khan SF, Gupta S, et al. Comparison of the conventional diagnosticmodalities, bactec culture and polymerase chain reaction test for diagnosis oftuberculosis [J]. Indian J Med Microbiol2005,23:29~33.
[8]谷海瀛.细菌形态学检查及其意义[J].中华检验医学杂志,2009,32(11):1315~1320
[9]中华人民共和国卫生部.中华人民共和国卫生行业标准——肺结核诊断标准(WS288—2008)[S],2008.
[10] Penedo MG, Carreira MJ, Mosquera A, et al. Computer Aided Diagnosis: ANeural Network Based Approach to Lung Nodule Detection [J]. IEEETransactions on Medical Imaging,1998,17(6):872~880
[11]李兵等.计算机图像处理技术在医学影像中的进展与应用[J].生物医学工程学杂志1993,10(4):360-363.
[12]庄天戈.计算机在生物医学中的应用[M].北京:科学出版社2000.
[13]黄波田.血细胞学检验[M].河南:河南医科大学出版社,1998
[14] Veropoulos K, Campbell C, Learmonth G, et al. The automatic identification oftubercle bacilli using image processing and neural computing techniques[C]. In:Proceedings of the eighth international conference on artificial neural networks,1998,2:797~802.
[15] Veropoulos K, Campbell C, Learmonth G, et al. The Automated identification oftubercle bacilli in sputum-a preliminary investigation [J]. Analytical andQuantitative Cytology and Histology,1999,21(4):277~281.
[16] Canny JF: A computational approach to edge detection [J]. IEEE Transactions onPattern Analysis and Machine Intelligence,1986,8:679~698.
[17] Shin C, Yun U, Kim H, et al. A Hybrid Approach of Neural Networks andMemory-based Learning to Data Mining [J]. IEEE Transactions on NeuralNetworks,2000,11(3):637~646
[18] Hopfield JJ. Neurons with Graded Response Have Collective ComputationalProperies Like Those of Two Sate Neurons. Proc.Natl.Acad.Sci.USA,1984,81(8):3088~3092
[19] Hart PE. The Condensed Nearest Neighbor rule [J]. IEEE Transaction onInformation Theory,1968,14(3):515~516
[20] Zhang B, Srihari SN. Fast K-Nearest Neighbor Classification UsingClustering-based Trees. IEEE Transactions on Pattern Analysis and MachineIntelligence,2004,26(4):525~528
[21] Forero M, Sierra E, Alvarez J, et al. Automatic sputum color image segmentationfor tuberculosis diagnosis[C].Proc. of the SPIE, San Diego,2001:251–261
[22] Forero M, Fuzzy thresholding and histogram analysis [J]. Fuzzy filters for imageprocessing. Series Studies in Fuzziness and Soft Computing, Springer Verlang,2002.
[23] Forero M, Sroubek F,Cristbal G, et al.Segmentation, autofocusing and signatureextraction of tuberculosis sputum images [C],SPIE Proc. Photonic Devices andAlgorithms for Computing IV, July,2002, Seattle, WA,4788.
[24] Forero M, Sroubek F,Cristbal G. Identification of tuberculosis bacteria based onshape and color, Real-Time Imaging [J].2004,10(4):251-262.
[25] Forero M, Sroubek F, Cristbal G. Automatic identification of Mycobacteriumtuberculosis by Gaussian Mixture Models[J].Journal of Microscopy,2006,223(2):120-132.
[26] Khutlang R, Krishnan S, Dendere R, et al. Classification of Mycobacteriumtuberculosis in images of ZN-stained sputum smears [J]. IEEE Transactions onInformation Technology in Biomedicine,2009,14(4):949-957
[27] Raof R A A., Salleh Z, Sahidan S I, et al. Color thresholding method for imagesegmentation algorithm of Ziehl-Neelsen sputum slide images[C].5thInternational Conference on Electrical Engineering, Computing Science andAutomatic Control (CCE2008), Mexico City,2008.
[28] Khutlang R, Krishnan S, Whitelaw A, et al. Detection of tuberculosis in sputumsmear images using two one-class classifiers[C]. IEEE International Symposiumon Biomedical Imaging: From Nano to Macro, ISBI'09.2009:1007~1010
[29] Khutlang R, Krishnan S, Whitelaw A, et al. Automated detection of tuberculosisin Ziehl-Neelsen-stained sputum smears using two one-class classifiers [J].Journal of Microscopy,2010,237(1):96~102.
[30] Makkapati V, Agrawal R, Acharya R. Segmentation and classification ofTuberculosis Bacilli from ZN-stained sputum smear images[C].5th Annual IEEEConference on Automation Science and Engineering Bangalore, India, August,2009:217-220
[31] Osman MK, Mashor MY, Jaafar H. Colour Image Enhancement using Bright andDark Stretching Techniques for Tissue based Tuberculosis Bacilli Detection[C].International Conference on Man-Machine System (ICoMMS). Batu Ferringhi,Penang, Malaysia, October,2009.
[32] Osman MK, Mashor MY, Jaafar H. Combining Threshoding and ClusteringTechniques for Mycobacterium tuberculosis Segmentation in Tissue Sections [J].Australian Journal of Basic and Applied Sciences,2011,5(12):1270~1279.
[33] Osman MK, Mashor MY, Jaafar H, et al. Segmentation of Tuberculosis Bacilli inZiehl-Neelsen-Stained Tissue Images based on K-Mean Clustering Procedure.Proceedings of the3rd International Conference on Intelligent&AdvancedSystems (ICIAS2010), Kuala Lumpur, Malaysia,2010.
[34] Osman MK, Mashor MY. Detection of Mycobacterium Tuberculosis in ZiehlNeelsen Stained Tissue Images using Zemike Moments and Hybrid MultilayeredPerceptron Network[C].2011.
[35] Osman MK, Mashor MY.Tuberculosis Bacilli Detection in Ziehl Neelsen StainedTissue using Affine Moment Invariants and Extreme Learning Machine[C]. IEEE7th International Collquium on Signal Processing and its Applications.2011.
[36] Huang G, Zhu Q and Siew C. Extreme learning machine: theory and applications[J].Neurocomputing.2006,70(1):489-501.
[37]于桂玲.基于计算机辅助的结核病检测及图像处理系统[D].北京:北京化工大学,2006.
[38]钟建军.结核菌动态显微图像智能识别关键技术研究[D].上海.东华大学,2009.
[39]孙涵璇.显微图像结核杆菌提取技术研究[D].北京:北京航空航天大学,2010.
[40] Yongping Zhai, Yunhui Liu, Dongxiang Zhou and Shun Liu.Automaticidentification of mycobacterium tuberculosis from ZN-stained sputum smear:Algorithm and system design [C].2010IEEE International Conference onRobotics and Biomimetics (ROBIO2010).Dec.2010, Tianjin, China,41-46.
[41] Yongping Zhai, Dongxiang Zhou, Shun Liu, Yunhui Liu and Wai-keungFung.Content based focus measure for robust autofocusing of microscopy inbiomedical applications[C]. IEEE4th International Conference on Nano/Molecular Medicine and Engineering. Dec,2010, HongKong, China,130-135.
[42]翟永平,刘云辉,周东翔,刘顺.稀疏图像内容情况下显微镜自动聚焦算法[J].软件学报,2012,23(5):1281~1294.
[43]翟永平,周东翔,刘云辉,刘顺,彭科举.聚焦函数性能评价指标设计及最优函数选取[J].光学学报,2011,31(4):0418002.
[44]翟永平,周东翔,刘云辉.极小景深条件下显微镜大范围聚焦算法[J].光学学报,2012,32(4):0418001.
[45] Wu Q, Merchant F.A and Castleman K.R.Microscope Image Processing[M],2008, Elsevier Inc.
[46] K.R.Spring,M.W.Davidson.Depth of Field and Depth of Focus.2010,http://www.microscopyu.com/
[47] Liron Y et al. Laser Autofocusing System for High-Resolution Cell BiologicalImaging [J].Journal of Microscopy.2006,221:145~151.
[48] Mendelsohn ML and Mayall BH. Computer-Oriented Analysis of HumanChromosomes-III focus [J]. Computers in Biology and Medicine.1972,2:137~150.
[49] Neill EL, Introduction to Statistical Optics [M]. Courier Dover Publications,2004.
[50] Goodman JW, Introduction to Fourier Optics Third Edition [M], Roberts andCompany,2005.
[51] ScottC, Introduction to Optics and Optical Imaging [M], IEEE Press,1998.
[52] Tenenbaum JM. Accommodation in computer vision [D]. Stanford: StanfordUniversity,1970.
[53] Jarvis R.A. Focus optimization criteria for computer image processing [J].Microscope,1976,24(2):163-180.
[54] Nayar SK, Nakagawa Y. Shape from focus [J]. IEEE Trans. on Pattern Analysisand Machine Intelligence,1994,16(8):824~831.
[55] Brenner JF, Dew BS, Horton JB, King T, Neurath PW, Selles WD. An automatedmicroscope for cytological research a preliminary evaluation [J]. Journal ofHistochemistry and Cytochemistry,1976,24(1):100~111
[56] Ge Y, Nelson BJ. Wavelet-Based autofocusing and unsupervised segmentation ofmicroscopic images[C]. In: Proc. of the IEEE/RSJ Int’l Conf. on IntelligentRobots and Systems.2003.21432148.
[57] Makkapati VV. Improved wavelet-based microscope autofocusing for bloodsmears by using segmentation[C]. In: Proc. of the5th Annual IEEE Conf. onAutomation Science and Engineering. Bangalore,2009.
[58]郭丙华,廖启亮,余志.基于小波变换的快速自动聚焦算法[J].中山大学学报(自然科学版),2007,46(2):1215.
[59] Materj K, Janez P, Matej P, et al. A Bayes-Spectral-Entropy-Based Measure ofCamera Focus Using a Discrete Cosine Transform [J]. Pattern RecognitionLetters.
[60] Subbarao M, Tyan JK. Selecting the optimal focus measure for autofocusing anddepth-from-focus [J]. IEEE Trans. on Pattern Analysis and Machine Intelligence,1998,20(8):864870.
[61] Santos A, Solórzano CO, Vaquero JJ, et al. Evaluation of autofocus functions inmolecular cytogenetic analysis [J]. Journal of Microscopy,1997,188(3):264272.
[62] Sun Y, Duthaler S, Nelson BJ. Autofocusing algorithm selection in computermicroscopy[C]. In: Proc. of the Intelligent Robots and Systems (IROS2005).2005.
[63] Russell MJ, Douglas TS. Evaluation of autofocus algorithms for tuberculosismicroscopy[C]. In: Proc. of the29th Annual Int’l Conf. of the IEEE EMBS CitéInternationale. Lyon,2007.
[64]孙杰,袁跃辉,王传永.数字图像处理自动图像聚焦算法的分析和比较[J].光学学报,2007,27(1):35～39.
[65] Luthi BS, Thomas N, Hwiid SF, Rueffer P. An efficient autofocus algorithm fora visible microscope on a Mars lander [J]. Planetary and Space Science,2010,58(10):1258~1264.
[66] Choi KS, Lee JS, Koo SJ. New autofocusing technique using the frequencyselective weighted median filter for video cameras [J]. IEEE Transactions onConsumer Electronics1999;45(3):820～827.
[67] Jie He, Rongzhen Zhou, Zhiliang Hong. Modified fast climbing searchauto-focus algorithm with adaptive step size searching technique for digitalcamera. IEEE Transactions on Consumer Electronics,2003;49:257～262.
[68] Krotkov E P. Active computer vision by cooperative focus and stereo [M]. NewYork: Springer-Verlag,1989.
[69]陈国金.数字图像自动聚焦技术研究及系统实现[D].西安:西安电子科技大学,2007.
[70] Pentland AP.A new sense for depth of filed [J] IEEE Transactions on PatternAnalysis and Machine Intelligence.1987,9(4):523~531.
[71] Subbarao M. Parallel Depth Recovery by Changing Camera Parameters[C]Second International Conference on ComputerVision, Florida, USA, December1988:149-155.
[72] Subbarao M and Natarajan G. Depth Recovery from Blurred Edges[C]Proceedings of the IEEE Conference on Computer Vision and PatternRecognition, Ann Arbor, Michigan, pp.498-503, June1988.
[73] Subbarao M and Surya G. Depth from Defocus: A Spatial Domain Approach.International Journal of Computer Vision,1994,13(3):271-294.
[74]李奇.数字自动对焦技术的理论及实现方法研究[D].浙江：浙江大学,2004
[75] M.E.Moghaddam.Out of Focus Blur Estimation Using Genetic Algorithm.Journal of Computer Science4(4):298-304,2008
[76] Oku H, Ishikawa M, et al. High-speed autofocusing of a cell using diffractionpattern[J], Optics Express.2006,14(9):3952~3960.
[77] Oku H, Makise S and Ishikawa M. Serial Algorithm for High speedAutofocusing of Cells using Depth From Diffraction(DFDi)Method.2008IEEEInternational Conference on Robotics and Automation, Pasadena, USA, May,2008:3124~3129
[78]孙即祥.数字图象处理[M].河北教育出版社.1991.
[79]唐宗明,张利,谢攀.一种基于能量和熵的自动聚焦算法[J].电子学报,2003,31(4):552555.
[80] Vollath D. Automatic focusing by correlative methods [J]. Journal of Microscopy1987,147(3):279-288.
[81] Groen F, Young IT, and Ligntart G. Acomparison of different focus functions foruse in autofocus algorithms [J]. Cytometry1985,6:81–91.
[82] Yeo T, Jayasooriah SO and Sinniah R. Autofocusing for tissue microscopy.Image Vis. Comput.1993,11:629–639.
[83] Chen GJ, Zhu MF, Wang YK et al. Study on definition evaluation fucntion basedon image contrast variation[C]. In: Proc the2007annual conference onInternational conference on computer engineering and applications, Australia,2007,254-258.
[84]李奇,冯华君,徐之海.自动对焦系统中图像非均匀采样的实验研究[J].光子学报,2003,32(12):14991501.
[85]朱孔凤,姜威,高赞,周贤,张进.自动聚焦系统中聚焦窗口的选择及参量的确定[J].光学学报,2006,26(6):836840.
[86]张乐,姜威,高赞.数字图像一阶矩的自动聚焦区域选择算法[J].光学技术,2008,34(2):163169.
[87] Herman B and Lemasters JJ. Optical Microscopy: Emerging Methods andApplications. Academic Press, New York,1993.
[88] Boddeke FR. Quantitative Fluorescence Microscopy [D]. Delf: Delf TechnischeUniversity,1999.
[89] McLanchlan G and Peel D. Finite Mixture Models. New York: John Wiley andSons,2000.
[90] Redner RA, Walker HF. Mixture density, maximum likelihood and the EMalgorithm [J]. SIAM Review,1984,26(2):195~239.
[91] McLanchlan G and Krishnan T. The EM algorithm and extensions. New York:John Wiley and Sons,1997.
[92] Cootes TF and Taylor CJ. Statistical Models of Appearance for Computer Vision
[R]. University of Manchester.2004. http://www.isbe.man.ac.uk/~bim/
[93] Cootes TF and Taylor CJ. Data driven refinement of active shape model search.In7th British Machine Vison Conference. Edinburgh, UK,1996:383~392.
[94] Cootes TF, Edwards GJ and Taylor CJ. Active appearance models. In5thEuropean Conference on Computer Vision, Springer, Berlin,1998,2:484~498.
[95] Cootes TF and Taylor CJ. Active shape models-their training and application [J].Computer Vision and Image Understanding,1995,61(1):38~59.
[96]王泽兵，杨朝晖.彩色图像分割技术研究[J].数字电视与数字视频.2005,(4):21~24
[97]林开颜,吴军辉,徐立鸿,彩色图像分割方法综述[J],中国图象图形学报,2005,10(1):1~10.
[98] Poynton CA. Frequently Asked Questions about Color[R].www.poynton.com.2005
[99] Wyszecki G, Stiles WS. Color science: concepts and methods, qualitative dataand formula.2ed. New York: Wiley,1982.
[100] Wladysslaw S and Andreas K. Colour Image Segmentation—A Survey[J].2004
[101]向日华,王润生.一种基于高斯混合模型的距离图像分割算法[J].软件学报.2003,14(7):1250~1257.
[102] Boor C．On Calculating with B-Spline[J]．Journal of Approx Theory.1972,(6):50~62．
[103]王增波，彭仁忠，宫兆刚. B样条曲线生成原理及实现[J].石河子大学学报(自然科学版).2009,27(1):118~121.
[104]王文涛,汪国昭.带形状参数的三角多项式均匀B样条[J].计算机学报.2005,28(7):1192~1198.
[105] Farin G. Curves and Surfaces for Computer Aided Geometric Design [M].NewYork: Academic Press,1988
[106] Barsky B.Computer Graohics and Geometric Modeling Using Beta-Splines [M].NewYork:Spring-Verlag,1988
[107] Edwards GJ, Taylor CJ, and Cootes TF. Interpreting face images using activeappearance models[C]. In Proc.3rd IEEE Int. Conf. on Automatic Face andGesture Recognition,1998.
[108]邓梁.基于ASM与AAM的人脸特征定位与匹配算法研究[D].长沙:中南大学,2009.
[109]山世光.人脸识别中若干关键问题的研究[D].北京:中国科学院计算技术研究所,2004.
[110]苏亚.主动表观模型的建模与匹配[D].西安:西安电子科技大学,2010.
[111] Gowe JC. Generalized Procrustes analysis. Psychometrika,1975,40:33~50.
[112] Chung DH and Sapiroy G.On the level lines and geometry of vector valuedimages [J]. IEEE SP Lett.,2000,7(9):241~243.
[113] Kreyszig E.Differential Geometry. University of Dorendo Press,1959.
[114]王大凯,侯榆青,彭进业.图像处理的偏微分方程方法[M].科学出版社.2008.
[115]许欣,陈强,孙怀江,夏德深.基于梯度域融合的图像视觉效果改善[J].中国图象图形学报.2011,16(2):278~286.
[116]陆金甫,关治.偏微分方程数值解法[M].北京:清华大学出版社,2004.
[117] Tschumperle D, Deriche R. Vector-valued image regularization with PDE's: acommon framework for different applications [J].IEEE Trans. on PatternAnalysis and Machine Intelligence.2005,27(4):506~517.
[118] Soille P. Morphological Image Analysis: Principles and Applications,2ed [M].Springer-Verlag,2008.
[119] Soille P.形态学图像分析原理与应用(第二版)[M].王小鹏等译.北京:清华大学出版社,2008.
[120] Rivest JF, Soille P and Beucher S. Morpholgical gradients [J]. Journal ofelectronic Imaging.1993,2(4):326~336.
[121] Persoon E and Fu K. Shape discrimination using Fourier descriptors [J]. IEEETrans Sys, Man, and Cyber.1977,7(3):170~178.
[122] Hu M. Visual pattern recognition by moment's invariant [J].IRE Trans.Inf.Theory, IT-8,1962:179~187.
[123] Flusser J and Suk T. Pattern recognition by affine moment invariants [J]. Patternrecognition,1993,26(1):167~174.
[124]冈萨雷斯.数字图像处理(第二版）[M].阮秋琦等译.北京:电子工业出版社,2003.
[125]章毓晋.中国图象工程[J].中国图象图形学报.2002,7(5):417~430.
[126]孙即祥.现代模式识别(第二版)[M].北京:高等教育出版社,2008.
[127] Hunt EB, Marin J, Stone PJ. Experiments in Introduction [M]. New York,Academic Press,1966.
[128] Quinlan JR．Introduction of decision trees[J]．Machine Learning, l.1986:81~106.
[129] Quinlan JR. Simplifying decision trees. International Journal of Man-MachineStudies,1987,27:221~234.
[130] Quinlan JR. C4.5Programs for machine learning [M].Morgan Kaufmann, SanMateo, Californie,1993.
[131] Breiman L, Friedman JH, Olshen RA, et al. Classification and regression trees.Statistics probability series. Wadsworth, Belmont,1984.
[132] Quinlan JR.Decision Trees and Multivalued Attributes [M]. MachineIntelligence. Oxford, England, Oxford Univ. Press.1988:305~318.
[133] Quinlan JR. Unknown attributes values in induction[C]. Proceedings of the SixthInternational Machine Learning Workshop Cornell. New York. MorganKaufmann,1989.
[134] Mingers J. An empirical comparison of pruning methods for decision-treeinduction. Machine Learning4,1989:227~243.
[135]黄泽宇.决策树分类器算法的研究[D].北京:北京交通大学,2006.
[136] Catlett J. Overpruning large decision trees. In proc.of the12th International JointConference on Artificial Intelligence(IJCAI91),Sydney, Australia,1991:764~769.
[137] Rokach L, Maimon O. Data minning and knowledge discovery handbook [M].
[138]王金玲.保持图像骨架连续的剪枝技术及相关问题研究[D].山东师范大学,2010.
[139] Blum H.Biological Shape and Visual Science [J].Journal of Theoretical Biology,1973,38:205-287.
[140] Blum H and Nagel RN. Shape Description Using Weighted Symmetric AxisFeatures [J].Pattern Recognition,1978,10(3):167–180.
[141] Manlandain G, Fernandez SV.Euclidean skeletons [J].Image and VisionComputing,1998,16:317-327.
[142] Lam L,Lee S,Suen Y．Thinning methodologies: a comprehensive survey[J].IEEETrans. on Pattern Analysis and Machine Intelligence.1992,14:869~885．
[143] Casasent D and Sturgill R. Optical hit-or-miss morphological transforms forATR [M]. In A. Tescher, editor. Application of Digital Image Processing,SPIE-1153:500~510.
[144] Zhang TY, Sue CY. A fast parallel algorithm for thinning digitaol patterns [J].Communications of the ACM.1984,27(3):236~239
[145]孙淑苍.有关医学研究论文中“灵敏度”和“特异度”指标使用的常见错误[J].中华劳动卫生职业病杂志.2007,25(3):181~182.
[146]李立明.流行病学(4版)[M].北京:人民卫生出版社,2001:50~58．
[2] World Health Organization.Global tuberculosis control-epidemiology, strategy,financing [M]. Geneva, Switzerland, WHO Press,2009.
[3]中华人民共和国卫生部.2011年度全国法定传染病疫情概况[EB/OL].www.moh.gov.cn,2012-02-10.
[4]中国疾病预防控制中心.中国结核病防治规划实施工作指南(2008年版)[S],2008.
[5]王会莲.结核细胞自动识别系统中关键技术的研究[D].北京:北京化工大学,2006.
[6] Bello AK, Njoku CH. Tuberculosis: current trends in diagnosis and treatment [J].Niger J Clin Pract.2005,8:118~124.
[7] Negi SS, Khan SF, Gupta S, et al. Comparison of the conventional diagnosticmodalities, bactec culture and polymerase chain reaction test for diagnosis oftuberculosis [J]. Indian J Med Microbiol2005,23:29~33.
[8]谷海瀛.细菌形态学检查及其意义[J].中华检验医学杂志,2009,32(11):1315~1320
[9]中华人民共和国卫生部.中华人民共和国卫生行业标准——肺结核诊断标准(WS288—2008)[S],2008.
[10] Penedo MG, Carreira MJ, Mosquera A, et al. Computer Aided Diagnosis: ANeural Network Based Approach to Lung Nodule Detection [J]. IEEETransactions on Medical Imaging,1998,17(6):872~880
[11]李兵等.计算机图像处理技术在医学影像中的进展与应用[J].生物医学工程学杂志1993,10(4):360-363.
[12]庄天戈.计算机在生物医学中的应用[M].北京:科学出版社2000.
[13]黄波田.血细胞学检验[M].河南:河南医科大学出版社,1998
[14] Veropoulos K, Campbell C, Learmonth G, et al. The automatic identification oftubercle bacilli using image processing and neural computing techniques[C]. In:Proceedings of the eighth international conference on artificial neural networks,1998,2:797~802.
[15] Veropoulos K, Campbell C, Learmonth G, et al. The Automated identification oftubercle bacilli in sputum-a preliminary investigation [J]. Analytical andQuantitative Cytology and Histology,1999,21(4):277~281.
[16] Canny JF: A computational approach to edge detection [J]. IEEE Transactions onPattern Analysis and Machine Intelligence,1986,8:679~698.
[17] Shin C, Yun U, Kim H, et al. A Hybrid Approach of Neural Networks andMemory-based Learning to Data Mining [J]. IEEE Transactions on NeuralNetworks,2000,11(3):637~646
[18] Hopfield JJ. Neurons with Graded Response Have Collective ComputationalProperies Like Those of Two Sate Neurons. Proc.Natl.Acad.Sci.USA,1984,81(8):3088~3092
[19] Hart PE. The Condensed Nearest Neighbor rule [J]. IEEE Transaction onInformation Theory,1968,14(3):515~516
[20] Zhang B, Srihari SN. Fast K-Nearest Neighbor Classification UsingClustering-based Trees. IEEE Transactions on Pattern Analysis and MachineIntelligence,2004,26(4):525~528
[21] Forero M, Sierra E, Alvarez J, et al. Automatic sputum color image segmentationfor tuberculosis diagnosis[C].Proc. of the SPIE, San Diego,2001:251–261
[22] Forero M, Fuzzy thresholding and histogram analysis [J]. Fuzzy filters for imageprocessing. Series Studies in Fuzziness and Soft Computing, Springer Verlang,2002.
[23] Forero M, Sroubek F,Cristbal G, et al.Segmentation, autofocusing and signatureextraction of tuberculosis sputum images [C],SPIE Proc. Photonic Devices andAlgorithms for Computing IV, July,2002, Seattle, WA,4788.
[24] Forero M, Sroubek F,Cristbal G. Identification of tuberculosis bacteria based onshape and color, Real-Time Imaging [J].2004,10(4):251-262.
[25] Forero M, Sroubek F, Cristbal G. Automatic identification of Mycobacteriumtuberculosis by Gaussian Mixture Models[J].Journal of Microscopy,2006,223(2):120-132.
[26] Khutlang R, Krishnan S, Dendere R, et al. Classification of Mycobacteriumtuberculosis in images of ZN-stained sputum smears [J]. IEEE Transactions onInformation Technology in Biomedicine,2009,14(4):949-957
[27] Raof R A A., Salleh Z, Sahidan S I, et al. Color thresholding method for imagesegmentation algorithm of Ziehl-Neelsen sputum slide images[C].5thInternational Conference on Electrical Engineering, Computing Science andAutomatic Control (CCE2008), Mexico City,2008.
[28] Khutlang R, Krishnan S, Whitelaw A, et al. Detection of tuberculosis in sputumsmear images using two one-class classifiers[C]. IEEE International Symposiumon Biomedical Imaging: From Nano to Macro, ISBI'09.2009:1007~1010
[29] Khutlang R, Krishnan S, Whitelaw A, et al. Automated detection of tuberculosisin Ziehl-Neelsen-stained sputum smears using two one-class classifiers [J].Journal of Microscopy,2010,237(1):96~102.
[30] Makkapati V, Agrawal R, Acharya R. Segmentation and classification ofTuberculosis Bacilli from ZN-stained sputum smear images[C].5th Annual IEEEConference on Automation Science and Engineering Bangalore, India, August,2009:217-220
[31] Osman MK, Mashor MY, Jaafar H. Colour Image Enhancement using Bright andDark Stretching Techniques for Tissue based Tuberculosis Bacilli Detection[C].International Conference on Man-Machine System (ICoMMS). Batu Ferringhi,Penang, Malaysia, October,2009.
[32] Osman MK, Mashor MY, Jaafar H. Combining Threshoding and ClusteringTechniques for Mycobacterium tuberculosis Segmentation in Tissue Sections [J].Australian Journal of Basic and Applied Sciences,2011,5(12):1270~1279.
[33] Osman MK, Mashor MY, Jaafar H, et al. Segmentation of Tuberculosis Bacilli inZiehl-Neelsen-Stained Tissue Images based on K-Mean Clustering Procedure.Proceedings of the3rd International Conference on Intelligent&AdvancedSystems (ICIAS2010), Kuala Lumpur, Malaysia,2010.
[34] Osman MK, Mashor MY. Detection of Mycobacterium Tuberculosis in ZiehlNeelsen Stained Tissue Images using Zemike Moments and Hybrid MultilayeredPerceptron Network[C].2011.
[35] Osman MK, Mashor MY.Tuberculosis Bacilli Detection in Ziehl Neelsen StainedTissue using Affine Moment Invariants and Extreme Learning Machine[C]. IEEE7th International Collquium on Signal Processing and its Applications.2011.
[36] Huang G, Zhu Q and Siew C. Extreme learning machine: theory and applications[J].Neurocomputing.2006,70(1):489-501.
[37]于桂玲.基于计算机辅助的结核病检测及图像处理系统[D].北京:北京化工大学,2006.
[38]钟建军.结核菌动态显微图像智能识别关键技术研究[D].上海.东华大学,2009.
[39]孙涵璇.显微图像结核杆菌提取技术研究[D].北京:北京航空航天大学,2010.
[40] Yongping Zhai, Yunhui Liu, Dongxiang Zhou and Shun Liu.Automaticidentification of mycobacterium tuberculosis from ZN-stained sputum smear:Algorithm and system design [C].2010IEEE International Conference onRobotics and Biomimetics (ROBIO2010).Dec.2010, Tianjin, China,41-46.
[41] Yongping Zhai, Dongxiang Zhou, Shun Liu, Yunhui Liu and Wai-keungFung.Content based focus measure for robust autofocusing of microscopy inbiomedical applications[C]. IEEE4th International Conference on Nano/Molecular Medicine and Engineering. Dec,2010, HongKong, China,130-135.
[42]翟永平,刘云辉,周东翔,刘顺.稀疏图像内容情况下显微镜自动聚焦算法[J].软件学报,2012,23(5):1281~1294.
[43]翟永平,周东翔,刘云辉,刘顺,彭科举.聚焦函数性能评价指标设计及最优函数选取[J].光学学报,2011,31(4):0418002.
[44]翟永平,周东翔,刘云辉.极小景深条件下显微镜大范围聚焦算法[J].光学学报,2012,32(4):0418001.
[45] Wu Q, Merchant F.A and Castleman K.R.Microscope Image Processing[M],2008, Elsevier Inc.
[46] K.R.Spring,M.W.Davidson.Depth of Field and Depth of Focus.2010,http://www.microscopyu.com/
[47] Liron Y et al. Laser Autofocusing System for High-Resolution Cell BiologicalImaging [J].Journal of Microscopy.2006,221:145~151.
[48] Mendelsohn ML and Mayall BH. Computer-Oriented Analysis of HumanChromosomes-III focus [J]. Computers in Biology and Medicine.1972,2:137~150.
[49] Neill EL, Introduction to Statistical Optics [M]. Courier Dover Publications,2004.
[50] Goodman JW, Introduction to Fourier Optics Third Edition [M], Roberts andCompany,2005.
[51] ScottC, Introduction to Optics and Optical Imaging [M], IEEE Press,1998.
[52] Tenenbaum JM. Accommodation in computer vision [D]. Stanford: StanfordUniversity,1970.
[53] Jarvis R.A. Focus optimization criteria for computer image processing [J].Microscope,1976,24(2):163-180.
[54] Nayar SK, Nakagawa Y. Shape from focus [J]. IEEE Trans. on Pattern Analysisand Machine Intelligence,1994,16(8):824~831.
[55] Brenner JF, Dew BS, Horton JB, King T, Neurath PW, Selles WD. An automatedmicroscope for cytological research a preliminary evaluation [J]. Journal ofHistochemistry and Cytochemistry,1976,24(1):100~111
[56] Ge Y, Nelson BJ. Wavelet-Based autofocusing and unsupervised segmentation ofmicroscopic images[C]. In: Proc. of the IEEE/RSJ Int’l Conf. on IntelligentRobots and Systems.2003.21432148.
[57] Makkapati VV. Improved wavelet-based microscope autofocusing for bloodsmears by using segmentation[C]. In: Proc. of the5th Annual IEEE Conf. onAutomation Science and Engineering. Bangalore,2009.
[58]郭丙华,廖启亮,余志.基于小波变换的快速自动聚焦算法[J].中山大学学报(自然科学版),2007,46(2):1215.
[59] Materj K, Janez P, Matej P, et al. A Bayes-Spectral-Entropy-Based Measure ofCamera Focus Using a Discrete Cosine Transform [J]. Pattern RecognitionLetters.
[60] Subbarao M, Tyan JK. Selecting the optimal focus measure for autofocusing anddepth-from-focus [J]. IEEE Trans. on Pattern Analysis and Machine Intelligence,1998,20(8):864870.
[61] Santos A, Solórzano CO, Vaquero JJ, et al. Evaluation of autofocus functions inmolecular cytogenetic analysis [J]. Journal of Microscopy,1997,188(3):264272.
[62] Sun Y, Duthaler S, Nelson BJ. Autofocusing algorithm selection in computermicroscopy[C]. In: Proc. of the Intelligent Robots and Systems (IROS2005).2005.
[63] Russell MJ, Douglas TS. Evaluation of autofocus algorithms for tuberculosismicroscopy[C]. In: Proc. of the29th Annual Int’l Conf. of the IEEE EMBS CitéInternationale. Lyon,2007.
[64]孙杰,袁跃辉,王传永.数字图像处理自动图像聚焦算法的分析和比较[J].光学学报,2007,27(1):35～39.
[65] Luthi BS, Thomas N, Hwiid SF, Rueffer P. An efficient autofocus algorithm fora visible microscope on a Mars lander [J]. Planetary and Space Science,2010,58(10):1258~1264.
[66] Choi KS, Lee JS, Koo SJ. New autofocusing technique using the frequencyselective weighted median filter for video cameras [J]. IEEE Transactions onConsumer Electronics1999;45(3):820～827.
[67] Jie He, Rongzhen Zhou, Zhiliang Hong. Modified fast climbing searchauto-focus algorithm with adaptive step size searching technique for digitalcamera. IEEE Transactions on Consumer Electronics,2003;49:257～262.
[68] Krotkov E P. Active computer vision by cooperative focus and stereo [M]. NewYork: Springer-Verlag,1989.
[69]陈国金.数字图像自动聚焦技术研究及系统实现[D].西安:西安电子科技大学,2007.
[70] Pentland AP.A new sense for depth of filed [J] IEEE Transactions on PatternAnalysis and Machine Intelligence.1987,9(4):523~531.
[71] Subbarao M. Parallel Depth Recovery by Changing Camera Parameters[C]Second International Conference on ComputerVision, Florida, USA, December1988:149-155.
[72] Subbarao M and Natarajan G. Depth Recovery from Blurred Edges[C]Proceedings of the IEEE Conference on Computer Vision and PatternRecognition, Ann Arbor, Michigan, pp.498-503, June1988.
[73] Subbarao M and Surya G. Depth from Defocus: A Spatial Domain Approach.International Journal of Computer Vision,1994,13(3):271-294.
[74]李奇.数字自动对焦技术的理论及实现方法研究[D].浙江：浙江大学,2004
[75] M.E.Moghaddam.Out of Focus Blur Estimation Using Genetic Algorithm.Journal of Computer Science4(4):298-304,2008
[76] Oku H, Ishikawa M, et al. High-speed autofocusing of a cell using diffractionpattern[J], Optics Express.2006,14(9):3952~3960.
[77] Oku H, Makise S and Ishikawa M. Serial Algorithm for High speedAutofocusing of Cells using Depth From Diffraction(DFDi)Method.2008IEEEInternational Conference on Robotics and Automation, Pasadena, USA, May,2008:3124~3129
[78]孙即祥.数字图象处理[M].河北教育出版社.1991.
[79]唐宗明,张利,谢攀.一种基于能量和熵的自动聚焦算法[J].电子学报,2003,31(4):552555.
[80] Vollath D. Automatic focusing by correlative methods [J]. Journal of Microscopy1987,147(3):279-288.
[81] Groen F, Young IT, and Ligntart G. Acomparison of different focus functions foruse in autofocus algorithms [J]. Cytometry1985,6:81–91.
[82] Yeo T, Jayasooriah SO and Sinniah R. Autofocusing for tissue microscopy.Image Vis. Comput.1993,11:629–639.
[83] Chen GJ, Zhu MF, Wang YK et al. Study on definition evaluation fucntion basedon image contrast variation[C]. In: Proc the2007annual conference onInternational conference on computer engineering and applications, Australia,2007,254-258.
[84]李奇,冯华君,徐之海.自动对焦系统中图像非均匀采样的实验研究[J].光子学报,2003,32(12):14991501.
[85]朱孔凤,姜威,高赞,周贤,张进.自动聚焦系统中聚焦窗口的选择及参量的确定[J].光学学报,2006,26(6):836840.
[86]张乐,姜威,高赞.数字图像一阶矩的自动聚焦区域选择算法[J].光学技术,2008,34(2):163169.
[87] Herman B and Lemasters JJ. Optical Microscopy: Emerging Methods andApplications. Academic Press, New York,1993.
[88] Boddeke FR. Quantitative Fluorescence Microscopy [D]. Delf: Delf TechnischeUniversity,1999.
[89] McLanchlan G and Peel D. Finite Mixture Models. New York: John Wiley andSons,2000.
[90] Redner RA, Walker HF. Mixture density, maximum likelihood and the EMalgorithm [J]. SIAM Review,1984,26(2):195~239.
[91] McLanchlan G and Krishnan T. The EM algorithm and extensions. New York:John Wiley and Sons,1997.
[92] Cootes TF and Taylor CJ. Statistical Models of Appearance for Computer Vision
[R]. University of Manchester.2004. http://www.isbe.man.ac.uk/~bim/
[93] Cootes TF and Taylor CJ. Data driven refinement of active shape model search.In7th British Machine Vison Conference. Edinburgh, UK,1996:383~392.
[94] Cootes TF, Edwards GJ and Taylor CJ. Active appearance models. In5thEuropean Conference on Computer Vision, Springer, Berlin,1998,2:484~498.
[95] Cootes TF and Taylor CJ. Active shape models-their training and application [J].Computer Vision and Image Understanding,1995,61(1):38~59.
[96]王泽兵，杨朝晖.彩色图像分割技术研究[J].数字电视与数字视频.2005,(4):21~24
[97]林开颜,吴军辉,徐立鸿,彩色图像分割方法综述[J],中国图象图形学报,2005,10(1):1~10.
[98] Poynton CA. Frequently Asked Questions about Color[R].www.poynton.com.2005
[99] Wyszecki G, Stiles WS. Color science: concepts and methods, qualitative dataand formula.2ed. New York: Wiley,1982.
[100] Wladysslaw S and Andreas K. Colour Image Segmentation—A Survey[J].2004
[101]向日华,王润生.一种基于高斯混合模型的距离图像分割算法[J].软件学报.2003,14(7):1250~1257.
[102] Boor C．On Calculating with B-Spline[J]．Journal of Approx Theory.1972,(6):50~62．
[103]王增波，彭仁忠，宫兆刚. B样条曲线生成原理及实现[J].石河子大学学报(自然科学版).2009,27(1):118~121.
[104]王文涛,汪国昭.带形状参数的三角多项式均匀B样条[J].计算机学报.2005,28(7):1192~1198.
[105] Farin G. Curves and Surfaces for Computer Aided Geometric Design [M].NewYork: Academic Press,1988
[106] Barsky B.Computer Graohics and Geometric Modeling Using Beta-Splines [M].NewYork:Spring-Verlag,1988
[107] Edwards GJ, Taylor CJ, and Cootes TF. Interpreting face images using activeappearance models[C]. In Proc.3rd IEEE Int. Conf. on Automatic Face andGesture Recognition,1998.
[108]邓梁.基于ASM与AAM的人脸特征定位与匹配算法研究[D].长沙:中南大学,2009.
[109]山世光.人脸识别中若干关键问题的研究[D].北京:中国科学院计算技术研究所,2004.
[110]苏亚.主动表观模型的建模与匹配[D].西安:西安电子科技大学,2010.
[111] Gowe JC. Generalized Procrustes analysis. Psychometrika,1975,40:33~50.
[112] Chung DH and Sapiroy G.On the level lines and geometry of vector valuedimages [J]. IEEE SP Lett.,2000,7(9):241~243.
[113] Kreyszig E.Differential Geometry. University of Dorendo Press,1959.
[114]王大凯,侯榆青,彭进业.图像处理的偏微分方程方法[M].科学出版社.2008.
[115]许欣,陈强,孙怀江,夏德深.基于梯度域融合的图像视觉效果改善[J].中国图象图形学报.2011,16(2):278~286.
[116]陆金甫,关治.偏微分方程数值解法[M].北京:清华大学出版社,2004.
[117] Tschumperle D, Deriche R. Vector-valued image regularization with PDE's: acommon framework for different applications [J].IEEE Trans. on PatternAnalysis and Machine Intelligence.2005,27(4):506~517.
[118] Soille P. Morphological Image Analysis: Principles and Applications,2ed [M].Springer-Verlag,2008.
[119] Soille P.形态学图像分析原理与应用(第二版)[M].王小鹏等译.北京:清华大学出版社,2008.
[120] Rivest JF, Soille P and Beucher S. Morpholgical gradients [J]. Journal ofelectronic Imaging.1993,2(4):326~336.
[121] Persoon E and Fu K. Shape discrimination using Fourier descriptors [J]. IEEETrans Sys, Man, and Cyber.1977,7(3):170~178.
[122] Hu M. Visual pattern recognition by moment's invariant [J].IRE Trans.Inf.Theory, IT-8,1962:179~187.
[123] Flusser J and Suk T. Pattern recognition by affine moment invariants [J]. Patternrecognition,1993,26(1):167~174.
[124]冈萨雷斯.数字图像处理(第二版）[M].阮秋琦等译.北京:电子工业出版社,2003.
[125]章毓晋.中国图象工程[J].中国图象图形学报.2002,7(5):417~430.
[126]孙即祥.现代模式识别(第二版)[M].北京:高等教育出版社,2008.
[127] Hunt EB, Marin J, Stone PJ. Experiments in Introduction [M]. New York,Academic Press,1966.
[128] Quinlan JR．Introduction of decision trees[J]．Machine Learning, l.1986:81~106.
[129] Quinlan JR. Simplifying decision trees. International Journal of Man-MachineStudies,1987,27:221~234.
[130] Quinlan JR. C4.5Programs for machine learning [M].Morgan Kaufmann, SanMateo, Californie,1993.
[131] Breiman L, Friedman JH, Olshen RA, et al. Classification and regression trees.Statistics probability series. Wadsworth, Belmont,1984.
[132] Quinlan JR.Decision Trees and Multivalued Attributes [M]. MachineIntelligence. Oxford, England, Oxford Univ. Press.1988:305~318.
[133] Quinlan JR. Unknown attributes values in induction[C]. Proceedings of the SixthInternational Machine Learning Workshop Cornell. New York. MorganKaufmann,1989.
[134] Mingers J. An empirical comparison of pruning methods for decision-treeinduction. Machine Learning4,1989:227~243.
[135]黄泽宇.决策树分类器算法的研究[D].北京:北京交通大学,2006.
[136] Catlett J. Overpruning large decision trees. In proc.of the12th International JointConference on Artificial Intelligence(IJCAI91),Sydney, Australia,1991:764~769.
[137] Rokach L, Maimon O. Data minning and knowledge discovery handbook [M].
[138]王金玲.保持图像骨架连续的剪枝技术及相关问题研究[D].山东师范大学,2010.
[139] Blum H.Biological Shape and Visual Science [J].Journal of Theoretical Biology,1973,38:205-287.
[140] Blum H and Nagel RN. Shape Description Using Weighted Symmetric AxisFeatures [J].Pattern Recognition,1978,10(3):167–180.
[141] Manlandain G, Fernandez SV.Euclidean skeletons [J].Image and VisionComputing,1998,16:317-327.
[142] Lam L,Lee S,Suen Y．Thinning methodologies: a comprehensive survey[J].IEEETrans. on Pattern Analysis and Machine Intelligence.1992,14:869~885．
[143] Casasent D and Sturgill R. Optical hit-or-miss morphological transforms forATR [M]. In A. Tescher, editor. Application of Digital Image Processing,SPIE-1153:500~510.
[144] Zhang TY, Sue CY. A fast parallel algorithm for thinning digitaol patterns [J].Communications of the ACM.1984,27(3):236~239
[145]孙淑苍.有关医学研究论文中“灵敏度”和“特异度”指标使用的常见错误[J].中华劳动卫生职业病杂志.2007,25(3):181~182.
[146]李立明.流行病学(4版)[M].北京:人民卫生出版社,2001:50~58．