基于支持向量机的医学图像分割
|
摘要
图像分割是图像处理领域广泛研究和应用的技术。随着实际应用要求的不断提高,图像分割的各种理论和方法也在不断拓展,图像分割应用的层面也在不断拓宽,其在医学图像分析中也得到了广泛应用。尽管目前的分割方法达到了一定效果,但是以往的分割技术多是基于传统统计学理论的方法,是基于样本数目趋于无穷大时的渐近理论,在对待高维特征、样本数较少等问题中很难获得好的效果,这些分割方法的推广能力也比较差。而实际问题多是样本数目有限的情况,因此基于上述理论的直观上很优秀的机器学习方法,在具体问题中的表现却可能达不到期望的效果。近年来出现的支持向量机方法建立在统计学习理论的VC维理论和结构风险最小化准则基础之上,基本原理是根据有限的样本信息在模型的复杂性和学习能力之间寻求最佳折衷,以期获得最好的推广能力。传统统计学习方法寻找的是样本数趋于无穷大时的最优值,而支持向量机方法得到的是有限样本信息下的最优解,其推广能力优于传统的学习方法。
本文主要对目前广泛研究的医学图像分割进行讨论,将支持向量机方法用于医学血细胞图像的分割。结合具体的实验分析和数据统计,说明了支持向量机核函数的选择、核参数的设置和惩罚因子的不同取值对医学图像分割性能的影响。同时研究了输入的样本特征信息,即窗口尺度对分割结果的影响。通过实验分析训练样本数目对分割结果的影响,验证了支持向量机方法的小样本应用特性。对加入噪声后的图像,描述了支持向量机方法的性能。将支持向量机分割方法与其他分割方法进行比较,显示了支持向量机分割方法的优势。支持向量机是一种描述机器学习问题的新方法,其特性还需进一步研究,其应用应向更多的领域推广后续的工作还很多。
Image segmentation is a widely studied and applied technology in image processing field. Along with the improvement of the actual application requirements, various theories and methods of image segmentation are constantly expanded, and the applications of image segmentation are more extensive. Image segmentation has been widely applied in medical image analysis. Although the current segmentation methods have achieved a certain effect, the previous segmentation techniques are mostly based on the traditional statistical methods, and they are based on the asymptotic theory when the number of the samples tends to infinity. In the problems of high dimensional feature and small sample number, they are so difficult to obtain good results and their generalization abilities are poor. But in practical applications, the number of the samples is often limited, so some very good methods of learning theory in the practical problems may not be satisfactory. Support vector machines which are emerged in recent years, are based on the theory of VC dimension of statistical learning theory and the principle of structural risk minimization. According to the limited information of the samples, support vector machines find the best compromise between model complexity and learning ability in order to obtain the best generalization ability. Traditional statistical learning methods find the optimal value when the number of the samples tends to infinity, while the support vector machines obtain the optimal solution under the limited information of the samples, so their generalization ability is better than the traditional learning methods.
In this paper, we mainly do some research on medical image segmentation, using the method of support vector machines to segment blood cells image. Combining the specific analysis of experiments and the statistics of data, we illustrate the effects which the kernel functions of support vector machines, the kernel parameters of kernel functions and the penalty factor C do to the performance of segmentation. After adding noise to the image, the performance of support vector machines has been described. At the same time, we study the impact of the characteristics of input samples on the segmentation results. The result of comparing the performance between support vector machines and other segmentation methods shows the advantage of support vector machines. Through the experiments of different numbers of training samples to the segmentation results, it has validated that the method based on support vector machines is better used in the field where the number of the samples is small. The method of support vector machines is a new method of machine learning, and their characteristics need to be studied further more. The application should be expanded to more areas, and there is more work follow-up to be done.
本文主要对目前广泛研究的医学图像分割进行讨论,将支持向量机方法用于医学血细胞图像的分割。结合具体的实验分析和数据统计,说明了支持向量机核函数的选择、核参数的设置和惩罚因子的不同取值对医学图像分割性能的影响。同时研究了输入的样本特征信息,即窗口尺度对分割结果的影响。通过实验分析训练样本数目对分割结果的影响,验证了支持向量机方法的小样本应用特性。对加入噪声后的图像,描述了支持向量机方法的性能。将支持向量机分割方法与其他分割方法进行比较,显示了支持向量机分割方法的优势。支持向量机是一种描述机器学习问题的新方法,其特性还需进一步研究,其应用应向更多的领域推广后续的工作还很多。
Image segmentation is a widely studied and applied technology in image processing field. Along with the improvement of the actual application requirements, various theories and methods of image segmentation are constantly expanded, and the applications of image segmentation are more extensive. Image segmentation has been widely applied in medical image analysis. Although the current segmentation methods have achieved a certain effect, the previous segmentation techniques are mostly based on the traditional statistical methods, and they are based on the asymptotic theory when the number of the samples tends to infinity. In the problems of high dimensional feature and small sample number, they are so difficult to obtain good results and their generalization abilities are poor. But in practical applications, the number of the samples is often limited, so some very good methods of learning theory in the practical problems may not be satisfactory. Support vector machines which are emerged in recent years, are based on the theory of VC dimension of statistical learning theory and the principle of structural risk minimization. According to the limited information of the samples, support vector machines find the best compromise between model complexity and learning ability in order to obtain the best generalization ability. Traditional statistical learning methods find the optimal value when the number of the samples tends to infinity, while the support vector machines obtain the optimal solution under the limited information of the samples, so their generalization ability is better than the traditional learning methods.
In this paper, we mainly do some research on medical image segmentation, using the method of support vector machines to segment blood cells image. Combining the specific analysis of experiments and the statistics of data, we illustrate the effects which the kernel functions of support vector machines, the kernel parameters of kernel functions and the penalty factor C do to the performance of segmentation. After adding noise to the image, the performance of support vector machines has been described. At the same time, we study the impact of the characteristics of input samples on the segmentation results. The result of comparing the performance between support vector machines and other segmentation methods shows the advantage of support vector machines. Through the experiments of different numbers of training samples to the segmentation results, it has validated that the method based on support vector machines is better used in the field where the number of the samples is small. The method of support vector machines is a new method of machine learning, and their characteristics need to be studied further more. The application should be expanded to more areas, and there is more work follow-up to be done.
引文
[1]杨强.支持向量机的模型及其在图像分割中的应用[D].重庆大学博士学位论文,2004.
[2]Nello Cristianini, John Shawe-Taylor著；李国正,王猛,曾华军译.支持向量机导论[M].北京：电子工业出版社,2004.3.
[3]Marina Kolesnik, Ales Fexa. How robust is the SVM wound segmentation [EB/OL]. http://www.fit.fraunhofer.de/-kolesnik/papers/pdf/norsig2006.pdf.
[4]章毓晋.图像处理和分析基础[M].北京：高等教育出版社,2003：180-203.
[5]王瑶.图像分割中关键技术的研究[D].江南大学硕十学位论文,2008.
[6]周道炳,朱卫纲.几种边缘检测算子的评估[J].指挥技术学院学报，2000,11(1)：59-63.
[7]Guido valli. Artificial neural networks for the segmentation of medical images [J].Technical Report No.940702, July,1994.
[8]Matthew C. Clark, Lawrence O. Hall, Dmitry B. Goldgof, et al. Automatic Tumor Segmentation Using knowledge-Based Technique [J]. IEEE Trans. Medical Imaging,1998, 17(2):187-201.
[9]刘洲峰,徐庆伟,李春雷.基于小波变换的图像分割研究[J].计算机应用与软件,2009,26(4)：62-64.
[10]宋雨潭,纪秀.基于数学形态学二值图像分割算法的研究[J].长春工程学院学报(自然科学版),2008,9(3)：68-70.
[11]徐海洋,曹万华,陈炜,郭丽艳.基于支持向量机方法的多目标图像分割[J].微电子学与计算机,2009,26(4)：5-11.
[12]薛志东,隋卫平,李利军.一种SVM与区域生长相结合的图像分割方法[J].计算机应用,2007,27(2)：.463-465.
[13]何伟,蒋加伏,齐琦.基于粗糙集理论和神经网络的图像分割方法[J].计算机工程与应用,2009,45(1)：188-190.
[14]孙敬飞,杨红卫.形态学分水岭算法在粘连大米图像分割中的应用[J].河南工业大学学报(自然科学版),2009,30(6)：84-87.
[15]关振红,周则明,何春,张云.基于PDE的卫星姿态图像分割[J].数据采集与处理,2009,24(B10)：132-135.
[16]蔡广宇,崔士林,吴昌林.弧焊机器人焊缝图像分割方法研究[J].计算机工程与应用,2008,44(11)：223-225.
[17]杨文柱,李道亮,魏新华等.棉花异性纤维图像分割方法[J].农业机械学报,2009,40(3)：156-160,171.
[18]胡宁,孟利民.离散傅里叶变换在指纹图像分割中的应用[J].计算机工程与应用,2009,45(26)：185-186,219.
[19]张炎强,吴晓红,何小海,王正勇.基于图论Isoperimetric算法的岩心图像分割[J].计算机工程与应用,2009,45(22)：218-220.
[20]C. Tsai, B. S. Manjunath, R. Jagadeesan. Automated segmentation of brain MR images [J].Pattern Recognition,1995,28(12):1825-1837.
[21]X. M. Pardo, P. Radeva, D. Cabello. Discriminant snakes for 3D reconstruction of anatomical organs [J].Medical Image Analysis,2003,7:293-310.
[22]朱长明,李晶等.谱聚类集成的淋巴结超声图像分割算法[J].计算机辅助设计与图形学
学报,2009,21(10)：1480-1486,1496.
[23]谢剑斌,刘通,李沛秦,程江华.一种基于优化区域OTSU的静脉图像分割算法[J].信号处理,2009,25(1)：87-89.
[24]黄力明.微粒群优化和区域生长结合的图像分割方法[J].计算机工程与应用.2009,45(28)：193-195,204.
[25]李志梅,肖德贵,王丽丽.基于小波变换和KFCM的彩色图像分割[J].计算机工程,2009,35(19)：203-205.
[26]James S. Duncan, Nicholas Ayache. Medical Image Analysis:Progress over Two Decades and the Challenges Ahead [J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2000,22(1):85-106.
[27]Dzung L. Pham, Chenyang Xu, Jerry L. Prince. A Survey of Current Methods in Medical Image Segmentation [C]. Technical Report JHU/ECE99-01, Johns HoPkins University,1998.
[28]Kevin P. Hinshaw. Shape-based models for interactive segmentation of medical images in Image Processing [J].Murray h.loew,editor, Proc.SPIE,1995,2434:771-780.
[29]W. M. Wells, W. E. L. Grimson, R. Kikins, et al. Adaptive Segmentation of MRI Data [J]. IEEE Trans. on Medical Imaging,1996,15(4):429-442.
[30]Jayaram K. Udupa, Supun Samarasekera. Fuzzy Connectedness and Object Definition: Theory, Algorithms, and Applications in Image Segmentation [J].Graphical Models and Image Processing,1996,58(3):246-241.
[31]Norberto Ezquerra, Rakesh Mullick. Knowledge-guided segmentation of 3D imagery [J]. Graphical Models and Image Processing,1996,58(6):510-523.
[32]田炜,周明全,耿国华.基于自组织特征映射神经网络的医学图像分割技术[J].计算机应用与软件,2008,25(1)：28-29,35.
[33]吕回,李峰,徐琼.基于小波包和改进的FCM的医学图像分割[J].计算机工程与应用,2008,44(12)：186-188.
[34]康晓东,何不廉,李志圣,张雪君.一种基于T-Snake模型的医学图像分割方法[J].计算机工程与应用,2008,44(1)：183-186.
[35]Sedelaar JPM, De la Rosette JJMCH, Beerlage HP, et al. Transrectal ultrasound imaging of the prostate:review and perspectives of recent developments [J].Prostate Cancer& Prostatic Diseases,1999,2(5):241.
[36]魏伟波,陈娅莎.基于支持向量机的AdaBoost人脸检测方法[J].计算机仿真,2008,25(6)：210-213.
[37]吴江,李太勇.基于加权支持向量机的金融时间序列预测[J].商业研究,2010,1：138-140.
[38]杨霞,宋顺林.基于加权近似支持向量机的文本分类研究[J].计算机工程与设计,2009,30(15)：3594-3596.
[39]吴进军,杜树新.SVM在车牌字符识别中的应用[J].电路与系统学报,2008,13(1)：84-87
[40]刘长良,孙晓娇,刘站营.支持向量机在锅炉飞灰含碳量软测量中的应用[J].电力科学与工程,2010,26(1)：39-43,59
[41]王德成,林辉.基于支持向量机的功率变换器开关管开路故障诊断[J].火力与指挥控制,2010,35(1)：126-128
[42]张翔,田金文,肖晓玲,柳健.支持向量机及其在医学图像分类中的应用[J].信号处理,2004,20(2)：208-212.
[43]边肇祺,张学工.模式识别(第二版)[M].北京：清华大学出版社,2006：284-304.
[44]V. Vapnik. Statistical learning theory [M]. New York:J. Wiley,1998.
[45]Kwang In Kim, Keechul Jung et al. Support vector machines for texture classification [J]. IEEE Trans. Pattern Analysis and Machine Intelligence,2002,24(11):1542-1550.
[46]Shutao Li, James T. Kwok, Hailong Zhua, Yaonan Wang. Texture classification using the support vector machines [J]. Pattern Recognition,36(2003):2883-2893.
[47]Dihua Li, Podolak I. T., Lee S. W.. Facial Component Extraction and Face Recognition with Support Vector Machines [A]. In Proc. Automatic Face and Gesture Recognition,2002: 76-81.
[48]Constantine Kotropoulos, Ioannis Pitas. Segmentation of ultrasonic images using Support Vector Machines [J]. Pattern Recognition Letters 24(2003):715-727.
[49]J. Miteran, S. Bouillant, E. Bourennane. SVM approximation for real-Time image segmentation by using an improved hyperrectangles-based method [J]. Real-Time Imaging, 9(2003):179-188.
[50]Daniel J. Strauss, Gabriele Steidl. Hybrid wavelet-support vector classification of waveforms [J]. Journal of Computational and Applied Mathematics,148(2002):375-400.
[51]Mingjun Song, Daniel Civco. Road Extraction Using SVM and Image Segmentation [J]. Photogrammetric Engineering and Remote Sensing,2004,70(12):1365-1372.
[52]黄建灯,陈庆全.基于支持向量机的红细胞彩色图像分割[J].计算机系统应用,2009,18(1)：35-38.
[53]徐芳.航空影像分割的支持向量机方法[D].武汉大学博士学位论文,2004.
[54]王瑞.基于支持向量机的医学图像处理[D].西安理工大学硕士学位论文,2008.
[55]Chih C. Chang, Chih J. Lin. LIBSVM:a Library for Support Vector Machines [EB/OL]. Department of Computer Science and Information Engineering, National TaiWan University. http://www.csie.ntu.edu.tw/-cjlin/libsvm/,2006.
[56]KR Muller, S. Mika, G. Ratsch, K. Tsuda, B. Scholkopf. An Introduction to Kernel-Based Learning Algorithms [J]. IEEE Transactions on Neural Networks,2001,12(2):181-201.
[57]Nobuyuki Otsu. A Threshold Selection Method from Gray-Level Histograms [J]. IEEE Trans. Sys., Man., Cyber,1979, SMC-9(1):62-66.
[58]史鑫.支持向量机在医学图像分割中的应用[D].首都医科大学硕十学位论文,2007.
[2]Nello Cristianini, John Shawe-Taylor著；李国正,王猛,曾华军译.支持向量机导论[M].北京：电子工业出版社,2004.3.
[3]Marina Kolesnik, Ales Fexa. How robust is the SVM wound segmentation [EB/OL]. http://www.fit.fraunhofer.de/-kolesnik/papers/pdf/norsig2006.pdf.
[4]章毓晋.图像处理和分析基础[M].北京：高等教育出版社,2003：180-203.
[5]王瑶.图像分割中关键技术的研究[D].江南大学硕十学位论文,2008.
[6]周道炳,朱卫纲.几种边缘检测算子的评估[J].指挥技术学院学报，2000,11(1)：59-63.
[7]Guido valli. Artificial neural networks for the segmentation of medical images [J].Technical Report No.940702, July,1994.
[8]Matthew C. Clark, Lawrence O. Hall, Dmitry B. Goldgof, et al. Automatic Tumor Segmentation Using knowledge-Based Technique [J]. IEEE Trans. Medical Imaging,1998, 17(2):187-201.
[9]刘洲峰,徐庆伟,李春雷.基于小波变换的图像分割研究[J].计算机应用与软件,2009,26(4)：62-64.
[10]宋雨潭,纪秀.基于数学形态学二值图像分割算法的研究[J].长春工程学院学报(自然科学版),2008,9(3)：68-70.
[11]徐海洋,曹万华,陈炜,郭丽艳.基于支持向量机方法的多目标图像分割[J].微电子学与计算机,2009,26(4)：5-11.
[12]薛志东,隋卫平,李利军.一种SVM与区域生长相结合的图像分割方法[J].计算机应用,2007,27(2)：.463-465.
[13]何伟,蒋加伏,齐琦.基于粗糙集理论和神经网络的图像分割方法[J].计算机工程与应用,2009,45(1)：188-190.
[14]孙敬飞,杨红卫.形态学分水岭算法在粘连大米图像分割中的应用[J].河南工业大学学报(自然科学版),2009,30(6)：84-87.
[15]关振红,周则明,何春,张云.基于PDE的卫星姿态图像分割[J].数据采集与处理,2009,24(B10)：132-135.
[16]蔡广宇,崔士林,吴昌林.弧焊机器人焊缝图像分割方法研究[J].计算机工程与应用,2008,44(11)：223-225.
[17]杨文柱,李道亮,魏新华等.棉花异性纤维图像分割方法[J].农业机械学报,2009,40(3)：156-160,171.
[18]胡宁,孟利民.离散傅里叶变换在指纹图像分割中的应用[J].计算机工程与应用,2009,45(26)：185-186,219.
[19]张炎强,吴晓红,何小海,王正勇.基于图论Isoperimetric算法的岩心图像分割[J].计算机工程与应用,2009,45(22)：218-220.
[20]C. Tsai, B. S. Manjunath, R. Jagadeesan. Automated segmentation of brain MR images [J].Pattern Recognition,1995,28(12):1825-1837.
[21]X. M. Pardo, P. Radeva, D. Cabello. Discriminant snakes for 3D reconstruction of anatomical organs [J].Medical Image Analysis,2003,7:293-310.
[22]朱长明,李晶等.谱聚类集成的淋巴结超声图像分割算法[J].计算机辅助设计与图形学
学报,2009,21(10)：1480-1486,1496.
[23]谢剑斌,刘通,李沛秦,程江华.一种基于优化区域OTSU的静脉图像分割算法[J].信号处理,2009,25(1)：87-89.
[24]黄力明.微粒群优化和区域生长结合的图像分割方法[J].计算机工程与应用.2009,45(28)：193-195,204.
[25]李志梅,肖德贵,王丽丽.基于小波变换和KFCM的彩色图像分割[J].计算机工程,2009,35(19)：203-205.
[26]James S. Duncan, Nicholas Ayache. Medical Image Analysis:Progress over Two Decades and the Challenges Ahead [J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2000,22(1):85-106.
[27]Dzung L. Pham, Chenyang Xu, Jerry L. Prince. A Survey of Current Methods in Medical Image Segmentation [C]. Technical Report JHU/ECE99-01, Johns HoPkins University,1998.
[28]Kevin P. Hinshaw. Shape-based models for interactive segmentation of medical images in Image Processing [J].Murray h.loew,editor, Proc.SPIE,1995,2434:771-780.
[29]W. M. Wells, W. E. L. Grimson, R. Kikins, et al. Adaptive Segmentation of MRI Data [J]. IEEE Trans. on Medical Imaging,1996,15(4):429-442.
[30]Jayaram K. Udupa, Supun Samarasekera. Fuzzy Connectedness and Object Definition: Theory, Algorithms, and Applications in Image Segmentation [J].Graphical Models and Image Processing,1996,58(3):246-241.
[31]Norberto Ezquerra, Rakesh Mullick. Knowledge-guided segmentation of 3D imagery [J]. Graphical Models and Image Processing,1996,58(6):510-523.
[32]田炜,周明全,耿国华.基于自组织特征映射神经网络的医学图像分割技术[J].计算机应用与软件,2008,25(1)：28-29,35.
[33]吕回,李峰,徐琼.基于小波包和改进的FCM的医学图像分割[J].计算机工程与应用,2008,44(12)：186-188.
[34]康晓东,何不廉,李志圣,张雪君.一种基于T-Snake模型的医学图像分割方法[J].计算机工程与应用,2008,44(1)：183-186.
[35]Sedelaar JPM, De la Rosette JJMCH, Beerlage HP, et al. Transrectal ultrasound imaging of the prostate:review and perspectives of recent developments [J].Prostate Cancer& Prostatic Diseases,1999,2(5):241.
[36]魏伟波,陈娅莎.基于支持向量机的AdaBoost人脸检测方法[J].计算机仿真,2008,25(6)：210-213.
[37]吴江,李太勇.基于加权支持向量机的金融时间序列预测[J].商业研究,2010,1：138-140.
[38]杨霞,宋顺林.基于加权近似支持向量机的文本分类研究[J].计算机工程与设计,2009,30(15)：3594-3596.
[39]吴进军,杜树新.SVM在车牌字符识别中的应用[J].电路与系统学报,2008,13(1)：84-87
[40]刘长良,孙晓娇,刘站营.支持向量机在锅炉飞灰含碳量软测量中的应用[J].电力科学与工程,2010,26(1)：39-43,59
[41]王德成,林辉.基于支持向量机的功率变换器开关管开路故障诊断[J].火力与指挥控制,2010,35(1)：126-128
[42]张翔,田金文,肖晓玲,柳健.支持向量机及其在医学图像分类中的应用[J].信号处理,2004,20(2)：208-212.
[43]边肇祺,张学工.模式识别(第二版)[M].北京：清华大学出版社,2006：284-304.
[44]V. Vapnik. Statistical learning theory [M]. New York:J. Wiley,1998.
[45]Kwang In Kim, Keechul Jung et al. Support vector machines for texture classification [J]. IEEE Trans. Pattern Analysis and Machine Intelligence,2002,24(11):1542-1550.
[46]Shutao Li, James T. Kwok, Hailong Zhua, Yaonan Wang. Texture classification using the support vector machines [J]. Pattern Recognition,36(2003):2883-2893.
[47]Dihua Li, Podolak I. T., Lee S. W.. Facial Component Extraction and Face Recognition with Support Vector Machines [A]. In Proc. Automatic Face and Gesture Recognition,2002: 76-81.
[48]Constantine Kotropoulos, Ioannis Pitas. Segmentation of ultrasonic images using Support Vector Machines [J]. Pattern Recognition Letters 24(2003):715-727.
[49]J. Miteran, S. Bouillant, E. Bourennane. SVM approximation for real-Time image segmentation by using an improved hyperrectangles-based method [J]. Real-Time Imaging, 9(2003):179-188.
[50]Daniel J. Strauss, Gabriele Steidl. Hybrid wavelet-support vector classification of waveforms [J]. Journal of Computational and Applied Mathematics,148(2002):375-400.
[51]Mingjun Song, Daniel Civco. Road Extraction Using SVM and Image Segmentation [J]. Photogrammetric Engineering and Remote Sensing,2004,70(12):1365-1372.
[52]黄建灯,陈庆全.基于支持向量机的红细胞彩色图像分割[J].计算机系统应用,2009,18(1)：35-38.
[53]徐芳.航空影像分割的支持向量机方法[D].武汉大学博士学位论文,2004.
[54]王瑞.基于支持向量机的医学图像处理[D].西安理工大学硕士学位论文,2008.
[55]Chih C. Chang, Chih J. Lin. LIBSVM:a Library for Support Vector Machines [EB/OL]. Department of Computer Science and Information Engineering, National TaiWan University. http://www.csie.ntu.edu.tw/-cjlin/libsvm/,2006.
[56]KR Muller, S. Mika, G. Ratsch, K. Tsuda, B. Scholkopf. An Introduction to Kernel-Based Learning Algorithms [J]. IEEE Transactions on Neural Networks,2001,12(2):181-201.
[57]Nobuyuki Otsu. A Threshold Selection Method from Gray-Level Histograms [J]. IEEE Trans. Sys., Man., Cyber,1979, SMC-9(1):62-66.
[58]史鑫.支持向量机在医学图像分割中的应用[D].首都医科大学硕十学位论文,2007.