医学显微细胞图像提取和分割技术的研究与实现
|
摘要
随着计算机图像技术的飞速发展,医学图像处理渐渐地在从人工识别向计算机自动识别的方向发展。医学显微细胞自动检测系统作为医学辅助软件,是图像技术和医学领域相结合的新兴产物。显微细胞自动检测系统能自动识别细胞图像中的细胞,然后进行形态学统计分析和判断,这可以大大节约诊断时间,并使远程诊断成为可能,所以开展显微细胞图像自动检测技术研究具有重大现实意义。
显微细胞图像自动检测系统一般分为图像预处理、图像细胞提取、粘连细胞分割、细胞统计及修正4个步骤。其中图像细胞提取、粘连细胞分割在细胞图像处理上都是技术难点,特别是在图像采集质量不佳,清晰度不高的情况下,如何有效提高细胞识别的正确率是近年来显微细胞图像处理技术的一个瓶颈。
图像噪声影响、图像提取效果不佳、粘连分割困难等诸多因素的约制。论文在最为关键的图像细胞提取和粘连细胞分割2个方面的技术上进行了改进。图像细胞提取即对图像中目标的提取过程。该部分首先介绍了论文所采用的细胞提取方法的理论基础,并分析了传统方法的局限性,在此基础上,提出了一种基于阈值处理的改进的目标提取算法;粘连细胞分割即对所提取的目标进行分割。该部分首先介绍了分水岭分割算法,并说明了其用于粘连细胞分割的优势所在。然后分析了传统分水岭分割方法处理细胞分割的缺点所在,在此基础上,提出了一种基于分水岭的改进粘连分割算法。实验结果表明,算法的改进取得了不错的效果。细胞信息统计及修正主要对细胞分割后进行统计。其中细胞信息统计包括细胞统计、细胞形态统计(面积、圆度、长宽比、矩形度)和细胞色度统计等;细胞修正主要介绍了孔洞填充技术,论文采用了细胞提取前和细胞提取后的双重填充方法,取得了不错的效果。
论文最后指出了在显微细胞图像提取和分割技术研究过程中的不足之处,在粘连分割和扫描算法上还有待改进。
With the rapid development of computer image technology, medical image management has changed from manually to cyber-detection gradually. As an accessorial medical software, medicinal micro-cell automatic detection system becomes a new product result from the combination of image technology and medical field. In clinic, medicinal micro-cell automatic detection system can identify the cells in the cell image and give the result quickly, so it saves a lot of time.
To pick up image characteristic is a difficulty all the time, especially when the image is illegible. How to improve this question is a bottleneck in image technology recently. The medicinal micro-cell automatic detection technology the thesis refered is also have some problems such as effect in image noise, image pick-up, overlapped dividing and so on.
The thesis gives up a auto-detection method based on other people. It divides the process with 4 steps, including image pretreatment, image cell pick-up, overlapped cell dividing, cell information count and fixture.
Image pretreatment means a series of work after open the image, including space filter enhancement, color space transition, gray result management and so on. The system uses median filter on space filter enhancement, and gives out the result after using median filter technology. It analyses the importance of space filter enhancement from compare the result from the image from using median filter technology to not using median filter technology.
Image cell pick-up means pick up the target in the image. This part introduces the basic theory of the technology of cells pick-up the thesis used first, and analyses the disadvantage of traditional technology. Then the thesis puts forward a new algorithm for picking up object based on threshold, and gives out the result in experiment.
Overlapped cell dividing means divide the target after picking up in the image. This part introduces watershed technology and the advantage in using watershed technology first. Then the thesis analyses the disadvantage of traditional watershed technology, and puts forward a new overlapped cell dividing method based on watershed. At the end of this part gives out the result in experiment.
Cell information count and fixture mainly calculate the information after cell dividing. Cell information count includes cell count, cell morphologic count, cell color count and so on. Cell fixture manly introduces hole fill technology. The thesis uses both fill technology before cell picking-up and after cell picking-up and have a good effect.
The thesis improves the two key steps, including image cell picking-up and overlapped cell dividing, in the whole system, and has a good effect. They are keystone of the thesis.
显微细胞图像自动检测系统一般分为图像预处理、图像细胞提取、粘连细胞分割、细胞统计及修正4个步骤。其中图像细胞提取、粘连细胞分割在细胞图像处理上都是技术难点,特别是在图像采集质量不佳,清晰度不高的情况下,如何有效提高细胞识别的正确率是近年来显微细胞图像处理技术的一个瓶颈。
图像噪声影响、图像提取效果不佳、粘连分割困难等诸多因素的约制。论文在最为关键的图像细胞提取和粘连细胞分割2个方面的技术上进行了改进。图像细胞提取即对图像中目标的提取过程。该部分首先介绍了论文所采用的细胞提取方法的理论基础,并分析了传统方法的局限性,在此基础上,提出了一种基于阈值处理的改进的目标提取算法;粘连细胞分割即对所提取的目标进行分割。该部分首先介绍了分水岭分割算法,并说明了其用于粘连细胞分割的优势所在。然后分析了传统分水岭分割方法处理细胞分割的缺点所在,在此基础上,提出了一种基于分水岭的改进粘连分割算法。实验结果表明,算法的改进取得了不错的效果。细胞信息统计及修正主要对细胞分割后进行统计。其中细胞信息统计包括细胞统计、细胞形态统计(面积、圆度、长宽比、矩形度)和细胞色度统计等;细胞修正主要介绍了孔洞填充技术,论文采用了细胞提取前和细胞提取后的双重填充方法,取得了不错的效果。
论文最后指出了在显微细胞图像提取和分割技术研究过程中的不足之处,在粘连分割和扫描算法上还有待改进。
With the rapid development of computer image technology, medical image management has changed from manually to cyber-detection gradually. As an accessorial medical software, medicinal micro-cell automatic detection system becomes a new product result from the combination of image technology and medical field. In clinic, medicinal micro-cell automatic detection system can identify the cells in the cell image and give the result quickly, so it saves a lot of time.
To pick up image characteristic is a difficulty all the time, especially when the image is illegible. How to improve this question is a bottleneck in image technology recently. The medicinal micro-cell automatic detection technology the thesis refered is also have some problems such as effect in image noise, image pick-up, overlapped dividing and so on.
The thesis gives up a auto-detection method based on other people. It divides the process with 4 steps, including image pretreatment, image cell pick-up, overlapped cell dividing, cell information count and fixture.
Image pretreatment means a series of work after open the image, including space filter enhancement, color space transition, gray result management and so on. The system uses median filter on space filter enhancement, and gives out the result after using median filter technology. It analyses the importance of space filter enhancement from compare the result from the image from using median filter technology to not using median filter technology.
Image cell pick-up means pick up the target in the image. This part introduces the basic theory of the technology of cells pick-up the thesis used first, and analyses the disadvantage of traditional technology. Then the thesis puts forward a new algorithm for picking up object based on threshold, and gives out the result in experiment.
Overlapped cell dividing means divide the target after picking up in the image. This part introduces watershed technology and the advantage in using watershed technology first. Then the thesis analyses the disadvantage of traditional watershed technology, and puts forward a new overlapped cell dividing method based on watershed. At the end of this part gives out the result in experiment.
Cell information count and fixture mainly calculate the information after cell dividing. Cell information count includes cell count, cell morphologic count, cell color count and so on. Cell fixture manly introduces hole fill technology. The thesis uses both fill technology before cell picking-up and after cell picking-up and have a good effect.
The thesis improves the two key steps, including image cell picking-up and overlapped cell dividing, in the whole system, and has a good effect. They are keystone of the thesis.
引文
[1]陈瑶(1998).角膜内皮细胞图像的自动检查系统的研究.《计算机工程与应用》. 5(1) : 40-41
[2]章毓晋(1999).图象处理与分析.清华大学出版社.
[3]章毓晋(2000).图象理解与计算机视觉.清华大学出版社.
[4] Van de Heijden.(1995). Edge and line Feature Extraction Based on covariance Models. IEEE Trans PAMI. 17(2):16-33
[5] Luo XP, Tian J, Zhu GY, et al(1999). A survey on image segmentation techniques. Pattern Recognition and Artificial Inteligence, 12(3): 300-312
[6] Lin Y, Tian J(2002). A survey on medical image segmentation methods. Pattern Recognition and Artificail Inteligence, 15(2):192-204.
[7] Cheng HD, Jiang XH, Sun Y, et al(2001). Color image segmentation: advances and prospects . Pattern Recognition, 34(12): 2259-2281.
[8] Ohta Y, Kanade T, Sakai T(1980). Color information for region segmentation. Computer Graphics and Image Processing, 13(3): 222-241.
[9] Nevatia(1977). A color edge detector and its use in scene segmentation. IEEE Transactions on Systems, Man and Cybernetics, 7(11): 820-826.
[10] Power W, Clist R(1996). Comparison of supervised learning techniques applied to color segmentation of fruit image. In:Proceeding of SPIE, Inteligent Roberts and Computer Vision. XV:Algorithms, Techniques, Active Vision, and Material Handing, pp 370-381.
[11] Hance G A, Umbaugh S E, Moss R H, et al(1996). Unsupervised color image segmentation with application to skin tumor b orders. IEEE Engineering in Medicine and Biology, 15(l):104-111.
[12] Pal N R, Pal S K(1993). A review on image segmentation techniques. Pattern Recognition, 26(9):1277-1294.
[13] Ohlander R, Price K, Reddy D R(1978). Pictures egmentation using a recursive region splitting method. Computer Graphics and Image Processing, 8(3):313-333.
[14] Underwood S A, Aggarwal J K(1977). Interactive computer analysis of aerial color infrared photographs. Computer Graphics and Image Processing, 6(1):1-24.
[15] Kurugollu F, Sankur B, Harmanci A E(2001). Color image segmentation using histogrammultithresholding and fusion.Image and Vision Computing, 19(13): 915-928.
[16] Ahuja N, Haralick R M, Rosenfeld A(1980). Neighbor gray levels as features in pixel classification. Pattern Recognition, 13(4): 251-260.
[17] Liu C Q, Cheng H(1995). An efficient clustering method for color image segmentation . Pattern Recognition and Artificial Intelligence, 8(A01):133-138.
[18] Carevic D, Caelli T(1997). Region-based coding of color image using Karhunen-Loeve Transform. Graphics Models and Image Processing, 59(1): 27-38.
[19] Wang X W,Shen L S, Wei B G, et al(2002). The focus segmentation of color ophthalmologic image based on modified K-means clustering and math ematical morphology. Chinese Journal of Biomedical Enginerring, 21(5):443-448.
[20] Bezdek J C(1981). Pattern recognition with fuzzy objective function algorithms. New York:Plenum Press.
[21] Ferri F, Vidal E(1992). Color image segmentation and labeling through multi-edit condensing.Pattern Recognition Letters, 13(8):561-568.
[22] Michael T U,Arbib A(1994). Color image segmentation using competitive learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(12): 1197-1206.
[23] Tremeau A, Borel N(1997). A region growing and merging algorithm to color segmentation. Pattern Recognition, 30(7):1191-1203.
[24] Vincent L, Soille P(1991). Watersheds in digitals paces:an efficient algorithm based on immersion simulations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(6):583-598.
[25] Geman S, Geman D(1984). Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE ransactions on Pattern Analysis and Machine Intelligence, 6(11):721-741.
[26] Huang C L, Cheng T Y,Chen C C(1992). Color images segmentation using scale space filter and Markov random field. Pattern Recognition, 25(10):1217-1229.
[27] Liu JQ , Yang YH(1994). Multiresolution color image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(7): 689-700.
[28] Tsai P, Chang C C, Hu Y C(2002). An adaptive two-stage edge detection scheme for digital color images[J]. Real-Time Imageing, 8(4):329-343.
[29] Tizhoosh H R(1997). Fuzzy image processing. Berlin:Springer-Verlag.
[30] Campadeli P, Medici D, Schettini R(1997). Color image segmentation using Hopfield networks.Image and Vision Computing, 15(3):161-166.
[31] Ong SH, Yea NC, Lee KH, et al(2002). Segmentation of color images using a two- stage self-organizing network.Image and Vision Computing, 20(4): 279-289.
[32] Papamarkos N, Strouthopoulous C, Andreadis I(2000). Multithresholding of color and gray-level images through a neural network technique. Image and Vision Computing, 18(3):213-222.
[33] Lescure P, Yedid V M, Dupoisot H, et al(1999). Color segmentation on biological microscope images. In:Proceeding of SPIE,Application of Artificial Neural Networks in Image Processing IV. San Jose, California, USA, 182-193.
[34] Li N, Li YF(2003). Feature encoding for unsupervised segmentation of color images. IEEE Transactions on systems, Man and Cybernetics, 33(3):438-447.
[35] Riseman E M, Arbib M A(1977). Computational techniques in the visual segmentation of static scenes. Computer Vision Graphics Image Processing, 6(3):221-276.
[36] Caelli T, Reye D(1993). On the classification of image regions by color,texture and shape. Pattern Recognition, 26(4):461-470.
[37] Valkealahti K, Oja E(1998). Reduced multidimensional histograms in color texture description. In:Proceeding of 1998 International Conference on Pattern Recognition. Brisbane, Australian, 1057-1061.
[38] Eom K B(1999). Segmentation of monochrome and colort exture using moving average modeling approach. Image Vision Computing, 17(3):233-244.
[39] Tao L M, Xu G Y(2001). Some color issues and application in machine vision. Chinese Science Bullet in, 46(3):178-190.
[40] Barnard K, Cardei V, Funt B(2002). A comparison of computational color constancy algorithms-part I: methodology and experiments with synthesized data. IEEE Transactions on Image Processing, 11(9):972-983.
[41] Barnard K, Cardei V, Funt B(2002). A comparison of computational color constancy algorithms-part I: experiments with image data. IEEE Transactions on Image Processing, 11(9):972-983.
[42] Tsin Y H, Collins R T, Ramesh V, et al(2001). Bayesian color constancy for outdoor object recognition. In:Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Kauai, Hawaii,USA, 8-14.
[43] Cast1eman K R(1998).数字图像处理.《电子工业出版社》, pp 318-325
[44] OTSU N(1979). A threshold selection method from gray-level histograms. IEEE Trans SMC, 9(1): 62-69.
[45]林开颜吴军辉徐立鸿(2005).彩色图像分割方法综述.《中国图象图形学报》, 10(1): 1-10..
[46] Rafael C(2000). Gonzalez Richard E. Woods. Digital Image Processing Second Edition.电子工业出版社.
[47]郭戈平西建胡敏周利莉(2004).基于迭代的癌细胞图象自动多阈值分割.《信息工程大学学报》, 5(3): 72-74
[48]刘关松吕嘉雯徐建国高敦岳(2002).一种新的二值图像标记的快速算法.《计算机工程与应用》, 03(4):57-59
[49] GOLDBERG D E(1989). Genetic algorithms in search optimization & machine learing. Addison Wesley, Reading MA.
[50]游迎容范影乐庞全(2005).基于距离变换的粘连细胞分割方法.《计算机工程与应用》, 20(3): 206-208
[51]李红松侯朝桢(2004).基于距离变换和形态学的顶点探测算法.《电子学报》, 32(1):38-41
[52]聂守平王鸣(2003).球状重叠区域图像分割与参数统计算法研究.《中国激光》, 30(11):1036-1040
[53]刘富强钱建生曹国清(2000).多媒体图像技术及应用.《人民邮电出版社》.2-4
[2]章毓晋(1999).图象处理与分析.清华大学出版社.
[3]章毓晋(2000).图象理解与计算机视觉.清华大学出版社.
[4] Van de Heijden.(1995). Edge and line Feature Extraction Based on covariance Models. IEEE Trans PAMI. 17(2):16-33
[5] Luo XP, Tian J, Zhu GY, et al(1999). A survey on image segmentation techniques. Pattern Recognition and Artificial Inteligence, 12(3): 300-312
[6] Lin Y, Tian J(2002). A survey on medical image segmentation methods. Pattern Recognition and Artificail Inteligence, 15(2):192-204.
[7] Cheng HD, Jiang XH, Sun Y, et al(2001). Color image segmentation: advances and prospects . Pattern Recognition, 34(12): 2259-2281.
[8] Ohta Y, Kanade T, Sakai T(1980). Color information for region segmentation. Computer Graphics and Image Processing, 13(3): 222-241.
[9] Nevatia(1977). A color edge detector and its use in scene segmentation. IEEE Transactions on Systems, Man and Cybernetics, 7(11): 820-826.
[10] Power W, Clist R(1996). Comparison of supervised learning techniques applied to color segmentation of fruit image. In:Proceeding of SPIE, Inteligent Roberts and Computer Vision. XV:Algorithms, Techniques, Active Vision, and Material Handing, pp 370-381.
[11] Hance G A, Umbaugh S E, Moss R H, et al(1996). Unsupervised color image segmentation with application to skin tumor b orders. IEEE Engineering in Medicine and Biology, 15(l):104-111.
[12] Pal N R, Pal S K(1993). A review on image segmentation techniques. Pattern Recognition, 26(9):1277-1294.
[13] Ohlander R, Price K, Reddy D R(1978). Pictures egmentation using a recursive region splitting method. Computer Graphics and Image Processing, 8(3):313-333.
[14] Underwood S A, Aggarwal J K(1977). Interactive computer analysis of aerial color infrared photographs. Computer Graphics and Image Processing, 6(1):1-24.
[15] Kurugollu F, Sankur B, Harmanci A E(2001). Color image segmentation using histogrammultithresholding and fusion.Image and Vision Computing, 19(13): 915-928.
[16] Ahuja N, Haralick R M, Rosenfeld A(1980). Neighbor gray levels as features in pixel classification. Pattern Recognition, 13(4): 251-260.
[17] Liu C Q, Cheng H(1995). An efficient clustering method for color image segmentation . Pattern Recognition and Artificial Intelligence, 8(A01):133-138.
[18] Carevic D, Caelli T(1997). Region-based coding of color image using Karhunen-Loeve Transform. Graphics Models and Image Processing, 59(1): 27-38.
[19] Wang X W,Shen L S, Wei B G, et al(2002). The focus segmentation of color ophthalmologic image based on modified K-means clustering and math ematical morphology. Chinese Journal of Biomedical Enginerring, 21(5):443-448.
[20] Bezdek J C(1981). Pattern recognition with fuzzy objective function algorithms. New York:Plenum Press.
[21] Ferri F, Vidal E(1992). Color image segmentation and labeling through multi-edit condensing.Pattern Recognition Letters, 13(8):561-568.
[22] Michael T U,Arbib A(1994). Color image segmentation using competitive learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(12): 1197-1206.
[23] Tremeau A, Borel N(1997). A region growing and merging algorithm to color segmentation. Pattern Recognition, 30(7):1191-1203.
[24] Vincent L, Soille P(1991). Watersheds in digitals paces:an efficient algorithm based on immersion simulations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(6):583-598.
[25] Geman S, Geman D(1984). Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE ransactions on Pattern Analysis and Machine Intelligence, 6(11):721-741.
[26] Huang C L, Cheng T Y,Chen C C(1992). Color images segmentation using scale space filter and Markov random field. Pattern Recognition, 25(10):1217-1229.
[27] Liu JQ , Yang YH(1994). Multiresolution color image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(7): 689-700.
[28] Tsai P, Chang C C, Hu Y C(2002). An adaptive two-stage edge detection scheme for digital color images[J]. Real-Time Imageing, 8(4):329-343.
[29] Tizhoosh H R(1997). Fuzzy image processing. Berlin:Springer-Verlag.
[30] Campadeli P, Medici D, Schettini R(1997). Color image segmentation using Hopfield networks.Image and Vision Computing, 15(3):161-166.
[31] Ong SH, Yea NC, Lee KH, et al(2002). Segmentation of color images using a two- stage self-organizing network.Image and Vision Computing, 20(4): 279-289.
[32] Papamarkos N, Strouthopoulous C, Andreadis I(2000). Multithresholding of color and gray-level images through a neural network technique. Image and Vision Computing, 18(3):213-222.
[33] Lescure P, Yedid V M, Dupoisot H, et al(1999). Color segmentation on biological microscope images. In:Proceeding of SPIE,Application of Artificial Neural Networks in Image Processing IV. San Jose, California, USA, 182-193.
[34] Li N, Li YF(2003). Feature encoding for unsupervised segmentation of color images. IEEE Transactions on systems, Man and Cybernetics, 33(3):438-447.
[35] Riseman E M, Arbib M A(1977). Computational techniques in the visual segmentation of static scenes. Computer Vision Graphics Image Processing, 6(3):221-276.
[36] Caelli T, Reye D(1993). On the classification of image regions by color,texture and shape. Pattern Recognition, 26(4):461-470.
[37] Valkealahti K, Oja E(1998). Reduced multidimensional histograms in color texture description. In:Proceeding of 1998 International Conference on Pattern Recognition. Brisbane, Australian, 1057-1061.
[38] Eom K B(1999). Segmentation of monochrome and colort exture using moving average modeling approach. Image Vision Computing, 17(3):233-244.
[39] Tao L M, Xu G Y(2001). Some color issues and application in machine vision. Chinese Science Bullet in, 46(3):178-190.
[40] Barnard K, Cardei V, Funt B(2002). A comparison of computational color constancy algorithms-part I: methodology and experiments with synthesized data. IEEE Transactions on Image Processing, 11(9):972-983.
[41] Barnard K, Cardei V, Funt B(2002). A comparison of computational color constancy algorithms-part I: experiments with image data. IEEE Transactions on Image Processing, 11(9):972-983.
[42] Tsin Y H, Collins R T, Ramesh V, et al(2001). Bayesian color constancy for outdoor object recognition. In:Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Kauai, Hawaii,USA, 8-14.
[43] Cast1eman K R(1998).数字图像处理.《电子工业出版社》, pp 318-325
[44] OTSU N(1979). A threshold selection method from gray-level histograms. IEEE Trans SMC, 9(1): 62-69.
[45]林开颜吴军辉徐立鸿(2005).彩色图像分割方法综述.《中国图象图形学报》, 10(1): 1-10..
[46] Rafael C(2000). Gonzalez Richard E. Woods. Digital Image Processing Second Edition.电子工业出版社.
[47]郭戈平西建胡敏周利莉(2004).基于迭代的癌细胞图象自动多阈值分割.《信息工程大学学报》, 5(3): 72-74
[48]刘关松吕嘉雯徐建国高敦岳(2002).一种新的二值图像标记的快速算法.《计算机工程与应用》, 03(4):57-59
[49] GOLDBERG D E(1989). Genetic algorithms in search optimization & machine learing. Addison Wesley, Reading MA.
[50]游迎容范影乐庞全(2005).基于距离变换的粘连细胞分割方法.《计算机工程与应用》, 20(3): 206-208
[51]李红松侯朝桢(2004).基于距离变换和形态学的顶点探测算法.《电子学报》, 32(1):38-41
[52]聂守平王鸣(2003).球状重叠区域图像分割与参数统计算法研究.《中国激光》, 30(11):1036-1040
[53]刘富强钱建生曹国清(2000).多媒体图像技术及应用.《人民邮电出版社》.2-4