红外图像分割的偏微分方程方法研究
|
摘要
图像分割是图像处理和计算机视觉领域中的一个基本问题,其目的是把图像分成不同的部分,每一部分灰度是同质的。目前为止,已经提出许多图像分割方法,其中,基于偏微分方程的方法被证明是一个十分有效的图像分割方法。红外技术的广泛应用,针对红外图像的分割,已经提出了许多经典的分割方法。由于红外图像本身的特性,红外图像分割的偏微分方程的方法正处于探索阶段,国内外学者也提出了一些自己的解决方案,但是效果不是很理想。
本文针对红外图像分割的偏微分方程方法存在的问题,做以下探索:
1)针对包含分水线这一类红外图像,本文提出了检测红外图像分水线的方法,在充分理论推导的基础上,通过实验验证了该方法的可行性。
2)深入研究了结构张量的理论,定义了一个新的刻画图像边缘的相干度量。并应用该相干度量,从图像灰度的变化角度,较好刻画了数字图像流线、T型、角点等边缘特征,较梯度模刻画的边缘更精确。在实验方面,给出了相干度量对可见光、红外、CT及人工合成图像的边缘刻画情况。在理论方面,同时也推导出了新相干度量是对梯度模刻画图像边缘特征的一种拓展。基于新的相干度量,提出了一个新的边缘停止函数,并以一个经典的偏微分方程分割方法为实验模型,以包括可见光图像、红外图像、B超图像在内的数字图像为测试图像,验证了新边缘停止函数具有较好的分割能力和抗噪性。
3)将提取红外图像分水线的形态-水平正投影最大值法与相干度量的PDE方法结合,验证了两种方法结合分割红外图像高效性。
Image segmentation is a fundamental problem in the field of image processing and computer vision. Its goal is to partition a given image into several dissimilar parts in each of which the intensity is homogeneous. So far, a wide variety of methods have been proposed to solve the image segmentation problem, in which partial differential equation (PDE) based methods have been proved to be an efficient framework for image segmentation.
The infrared image segmentation is widely used in precision weapons, civilian navigation, tracking and other fields. Because of its military and civilian aspects of the application, many of infrared image segmentation methods have been proposed, in which PDE-based methods have not been proved to be an efficient method for infrared image segmentation.
This dissertation focuses on PDE-based methods for infrared image segmentation; the main results are summarized as follows:
1. A morphological filtering and gradient operator for infrared image segmentation is presented to detect the sea-sky-line in the infrared image (sea-sky-line method).
2. All of edge-based PDE methods rely on the edge stopping function. It is typically a decreasing function of the gradient magnitude of Gaussian smoothed image. In identifying object edges, however, image gradient does not take into account some important features such as junctions and corners; this results in the inaccurate location of edges or even false segmentations. Based on a measure of the local coherence of image structure tensor, this paper proposes a new edge stopping function. Experimental results show that a PDE method using the new edge stopping function can significantly perform better in the location of edges, while it is much faster and more robustness to noise than the original method.
3. The above sea-sky-line method and PDE method with local coherence are applied to infrared image segmentation. Numerical results show the effectiveness and reliability of the proposed method.
本文针对红外图像分割的偏微分方程方法存在的问题,做以下探索:
1)针对包含分水线这一类红外图像,本文提出了检测红外图像分水线的方法,在充分理论推导的基础上,通过实验验证了该方法的可行性。
2)深入研究了结构张量的理论,定义了一个新的刻画图像边缘的相干度量。并应用该相干度量,从图像灰度的变化角度,较好刻画了数字图像流线、T型、角点等边缘特征,较梯度模刻画的边缘更精确。在实验方面,给出了相干度量对可见光、红外、CT及人工合成图像的边缘刻画情况。在理论方面,同时也推导出了新相干度量是对梯度模刻画图像边缘特征的一种拓展。基于新的相干度量,提出了一个新的边缘停止函数,并以一个经典的偏微分方程分割方法为实验模型,以包括可见光图像、红外图像、B超图像在内的数字图像为测试图像,验证了新边缘停止函数具有较好的分割能力和抗噪性。
3)将提取红外图像分水线的形态-水平正投影最大值法与相干度量的PDE方法结合,验证了两种方法结合分割红外图像高效性。
Image segmentation is a fundamental problem in the field of image processing and computer vision. Its goal is to partition a given image into several dissimilar parts in each of which the intensity is homogeneous. So far, a wide variety of methods have been proposed to solve the image segmentation problem, in which partial differential equation (PDE) based methods have been proved to be an efficient framework for image segmentation.
The infrared image segmentation is widely used in precision weapons, civilian navigation, tracking and other fields. Because of its military and civilian aspects of the application, many of infrared image segmentation methods have been proposed, in which PDE-based methods have not been proved to be an efficient method for infrared image segmentation.
This dissertation focuses on PDE-based methods for infrared image segmentation; the main results are summarized as follows:
1. A morphological filtering and gradient operator for infrared image segmentation is presented to detect the sea-sky-line in the infrared image (sea-sky-line method).
2. All of edge-based PDE methods rely on the edge stopping function. It is typically a decreasing function of the gradient magnitude of Gaussian smoothed image. In identifying object edges, however, image gradient does not take into account some important features such as junctions and corners; this results in the inaccurate location of edges or even false segmentations. Based on a measure of the local coherence of image structure tensor, this paper proposes a new edge stopping function. Experimental results show that a PDE method using the new edge stopping function can significantly perform better in the location of edges, while it is much faster and more robustness to noise than the original method.
3. The above sea-sky-line method and PDE method with local coherence are applied to infrared image segmentation. Numerical results show the effectiveness and reliability of the proposed method.
引文
[1] El-Sheikh.Performance investigation of a homing guided missile with positioning the seeker antenna[J]. Egyptian Army,2004, 20: 1-77.
[2] Kazufumi Kaneda,Eihachiro Nakamae,Eiji Takahashi,Kenichi Yazawa.An unmannedwatching system using video cameras[J].IEEE Computer Applications in Power,1990,4:20~24.
[3] Kai Wang,Yan Liu,Xiao Wei Sun.Small Moving Infrared Target Detection Algorithm under Low SNR Background[J].Dept of Electr&Electron.2009.8 61-7.
[4]范金荣. 21世纪前20年精确制导技术发展预测[J] .现代防御技术, 2003(1).
[5]张中南.发展中的红外成像制导技术[J ] .飞航导弹,2006 (1) :40242.
[6] Marshall C A.Quantitative and imaging performance of uncooled microbolometer sensor[J]. SPIE,1997,3061:191–197.
[7]刘建军.长江航行安全问题的研究[J].中国安全科学学报,2003,13(4):56-57.
[8]戴彤宇,聂武,刘伟力.长江干线船撞桥事故分析[J].中国航海,2002,53(4):89-91.
[9]杨爱新.长江内河运输船舶事故的探讨[J].南通航运职业技术学院学报,2006, 5(1):102-105.
[10]中华人民共和国交通部.2006年公路水路交通行业发展统计公报[Z].
[11]记红.红外技术基础与应用[M].北京:科学出版社,1979.
[12] Ershow M,Liu H.C,Perera A.GU.etc. Optical interference and nonlinearities inquantum-well infrared photodetectors[J].Physica E.2000,7(1-2):115~118.
[13]何乃甩,黄席樾,刘俊,权循宝.基于形态学重构的内河红外船舶目标检测[J].红外技术,2007,(7):419-421.
[14]雷选华,王江安,李树山.海空背景下红外点日标检测算法[J].激光与红外.Feb2001,3:32-34.
[15]刘靳.红外图像预处理及弱小目标检测方法研究[D].西安:西安电子科技大学2006.
[16]陈玉丹,周冰.红外小目标检测中的背景预处理技术研究[J].科学技术与工程,2006,Vol.6,No.18,Sep.2897~2899.
[17]刘松涛,周晓东,王成刚.复杂海空背景下鲁棒的海天线检测算法研究[J].光电工程,2006,33(8):5-10.
[18]杜奇,向健勇,袁胜春.基于边缘强度的红外图像阈值分割方法研究[J].红外与激光工程,2004,vol 33,pp:288-291.
[19]周赟,李久贤,夏良正.基于区域增长的红外图像分割[J].南京理工大学学报,Dec.2002,Vol.26 Supp.
[20] N Otsu.A Threshold Selection Method from Gray-Level Histograms[J]. IEEE TransSystem,Man and Cybernetics.1979,9(1).
[21] V Caselles, R Kimmel, G Sapiro. Geodesic active contours. International Journal of Computer Vision[J].1997,22(1):61.
[22] M. Li, C. He, and Yi Zhan. Adaptive level-set evolution without initial contours for image segmentation[EB/OL]. J. Electron. Imaging 20, 023004 (2011), DOI: 10.1117 / 1.3574770.
[23] C..Collet P.Thoure1. Active Contour Models for Infrared Cloudy Shapes Segmentation[J]. OCEANS '94. 'Oceans Engineering for Today's Technology and Tomorrow's Preservation.' 1994 :II/444 - II/448 vol.2.
[24] F. Bunyak, K. Palaniappan, S. K. Nath. Geodesic Active Contour Based Fusion of Visible and Infrared Video forPersistent Object Tracking[J]. Applications of Compute Vision. 2007, 6:17-22.
[25] T F Chan, L A Vese. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2): 266-277.
[26] Li, C Kao, C Gore, Z Ding. Implicit active contours driven by local binary fitting energy[J]. IEEE Computer Vision and Pattern Recognition, 2007, 6:17-22.
[27] Osher S, Sethian J A. Fronts propagating with curvature- dependent speed: algorithms based on Hamilton-Jacobi formulation [J]. Computational Physics, 1988, 79: 12-49.
[28] V. Caselles, F. Catte, T. Coll, F. Dibos, A geometric model for active contours in image processing [J], 1993, Numer Math. 66(1): 1-31.
[29] J.W. Lu, J.C. Ren, Y. Lu, et al. A Modified Canny Algorithm for Detecting Sky-Sea Line in Infrared Images[C]. Sixth International Conference on Intelligent Systems Design and Applications, 2006.pp. 289-294.
[30]章毓晋.图像工程(上)[M].北京.清华大学出版社.2006.3.
[31]张锋,杨树谦,倪汉昌.舰船红外图像特征提取及目标识别技术探讨[J].红外与激光技术,1991,20(2):21-25.
[32] Serra J.Image Analysis and Mathematical Morphology[J].New York:Academic Press,1982.
[33] Zhong Ji, Yuting Su, Jian Wang, Rui Hua.Robust sea-sky-line detection based onhorizontal projection and hough transformation[J]. Image and Signal Processing, 2009.1.
[34] Weickert J. Anisotropic Diffusion in Image Processing [M]. ECMI Series, Teubner-Verlag, Stuttgart, Germany, 1998.
[35] Clark M,Bobik A C,Geisler W S. Multi-channel texture analysis using localized spatial filter[J]. IEEE,PAMI,1990,12(1):55~73.
[36] Sapiro G.. Geometric Partial Differential Equations and Image Analysis [M]. Cambridge University Press, 1999.
[37] Aubert G, Kornprobst P. Mathematical Problems in Image Processing [M]. New York:Springer-Verlag. 2002.
[38] Caselles V, Kimmel R, and Sapiro G. Geodesic active contours [J]. Int J Comput Vison, 1997, 22(1):61-79.
[39] Li C, Xu C, Gui C, Fox M D. Level set evolution without re-initialization: a new variational formulation [C]. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005, 1: 430-436.
[40]余瑞星,朱冰,吕梅柏.一种新的水平集停止项函数选取方法研究[J].系统仿真学报, 2008, 20(22): 6154-6157.
[41]何传江,田巧玉.几何活动轮廓模型中停止速度函数的尺度变换[J].计算机工程与应用, 2008, 44(8): 82-84.
[42]何传江,唐利明.几何活动轮廓模型中停止速度场的异性扩散[J].软件学报, 2007, 18(2): 600-607.
[43] Yang C, Zheng S, Ye J. Level set contour extraction method based on support value filter [J]. Applied Mathematics and Computation, 2008, 205: 688-696.
[44]冯玉玲,何传江,李梦.不用高斯平滑的边缘活动轮廓模型[J].计算机工程与应用, 2010, 46 (36): 192-194.
[45]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,2001.29～35.
[2] Kazufumi Kaneda,Eihachiro Nakamae,Eiji Takahashi,Kenichi Yazawa.An unmannedwatching system using video cameras[J].IEEE Computer Applications in Power,1990,4:20~24.
[3] Kai Wang,Yan Liu,Xiao Wei Sun.Small Moving Infrared Target Detection Algorithm under Low SNR Background[J].Dept of Electr&Electron.2009.8 61-7.
[4]范金荣. 21世纪前20年精确制导技术发展预测[J] .现代防御技术, 2003(1).
[5]张中南.发展中的红外成像制导技术[J ] .飞航导弹,2006 (1) :40242.
[6] Marshall C A.Quantitative and imaging performance of uncooled microbolometer sensor[J]. SPIE,1997,3061:191–197.
[7]刘建军.长江航行安全问题的研究[J].中国安全科学学报,2003,13(4):56-57.
[8]戴彤宇,聂武,刘伟力.长江干线船撞桥事故分析[J].中国航海,2002,53(4):89-91.
[9]杨爱新.长江内河运输船舶事故的探讨[J].南通航运职业技术学院学报,2006, 5(1):102-105.
[10]中华人民共和国交通部.2006年公路水路交通行业发展统计公报[Z].
[11]记红.红外技术基础与应用[M].北京:科学出版社,1979.
[12] Ershow M,Liu H.C,Perera A.GU.etc. Optical interference and nonlinearities inquantum-well infrared photodetectors[J].Physica E.2000,7(1-2):115~118.
[13]何乃甩,黄席樾,刘俊,权循宝.基于形态学重构的内河红外船舶目标检测[J].红外技术,2007,(7):419-421.
[14]雷选华,王江安,李树山.海空背景下红外点日标检测算法[J].激光与红外.Feb2001,3:32-34.
[15]刘靳.红外图像预处理及弱小目标检测方法研究[D].西安:西安电子科技大学2006.
[16]陈玉丹,周冰.红外小目标检测中的背景预处理技术研究[J].科学技术与工程,2006,Vol.6,No.18,Sep.2897~2899.
[17]刘松涛,周晓东,王成刚.复杂海空背景下鲁棒的海天线检测算法研究[J].光电工程,2006,33(8):5-10.
[18]杜奇,向健勇,袁胜春.基于边缘强度的红外图像阈值分割方法研究[J].红外与激光工程,2004,vol 33,pp:288-291.
[19]周赟,李久贤,夏良正.基于区域增长的红外图像分割[J].南京理工大学学报,Dec.2002,Vol.26 Supp.
[20] N Otsu.A Threshold Selection Method from Gray-Level Histograms[J]. IEEE TransSystem,Man and Cybernetics.1979,9(1).
[21] V Caselles, R Kimmel, G Sapiro. Geodesic active contours. International Journal of Computer Vision[J].1997,22(1):61.
[22] M. Li, C. He, and Yi Zhan. Adaptive level-set evolution without initial contours for image segmentation[EB/OL]. J. Electron. Imaging 20, 023004 (2011), DOI: 10.1117 / 1.3574770.
[23] C..Collet P.Thoure1. Active Contour Models for Infrared Cloudy Shapes Segmentation[J]. OCEANS '94. 'Oceans Engineering for Today's Technology and Tomorrow's Preservation.' 1994 :II/444 - II/448 vol.2.
[24] F. Bunyak, K. Palaniappan, S. K. Nath. Geodesic Active Contour Based Fusion of Visible and Infrared Video forPersistent Object Tracking[J]. Applications of Compute Vision. 2007, 6:17-22.
[25] T F Chan, L A Vese. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2): 266-277.
[26] Li, C Kao, C Gore, Z Ding. Implicit active contours driven by local binary fitting energy[J]. IEEE Computer Vision and Pattern Recognition, 2007, 6:17-22.
[27] Osher S, Sethian J A. Fronts propagating with curvature- dependent speed: algorithms based on Hamilton-Jacobi formulation [J]. Computational Physics, 1988, 79: 12-49.
[28] V. Caselles, F. Catte, T. Coll, F. Dibos, A geometric model for active contours in image processing [J], 1993, Numer Math. 66(1): 1-31.
[29] J.W. Lu, J.C. Ren, Y. Lu, et al. A Modified Canny Algorithm for Detecting Sky-Sea Line in Infrared Images[C]. Sixth International Conference on Intelligent Systems Design and Applications, 2006.pp. 289-294.
[30]章毓晋.图像工程(上)[M].北京.清华大学出版社.2006.3.
[31]张锋,杨树谦,倪汉昌.舰船红外图像特征提取及目标识别技术探讨[J].红外与激光技术,1991,20(2):21-25.
[32] Serra J.Image Analysis and Mathematical Morphology[J].New York:Academic Press,1982.
[33] Zhong Ji, Yuting Su, Jian Wang, Rui Hua.Robust sea-sky-line detection based onhorizontal projection and hough transformation[J]. Image and Signal Processing, 2009.1.
[34] Weickert J. Anisotropic Diffusion in Image Processing [M]. ECMI Series, Teubner-Verlag, Stuttgart, Germany, 1998.
[35] Clark M,Bobik A C,Geisler W S. Multi-channel texture analysis using localized spatial filter[J]. IEEE,PAMI,1990,12(1):55~73.
[36] Sapiro G.. Geometric Partial Differential Equations and Image Analysis [M]. Cambridge University Press, 1999.
[37] Aubert G, Kornprobst P. Mathematical Problems in Image Processing [M]. New York:Springer-Verlag. 2002.
[38] Caselles V, Kimmel R, and Sapiro G. Geodesic active contours [J]. Int J Comput Vison, 1997, 22(1):61-79.
[39] Li C, Xu C, Gui C, Fox M D. Level set evolution without re-initialization: a new variational formulation [C]. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005, 1: 430-436.
[40]余瑞星,朱冰,吕梅柏.一种新的水平集停止项函数选取方法研究[J].系统仿真学报, 2008, 20(22): 6154-6157.
[41]何传江,田巧玉.几何活动轮廓模型中停止速度函数的尺度变换[J].计算机工程与应用, 2008, 44(8): 82-84.
[42]何传江,唐利明.几何活动轮廓模型中停止速度场的异性扩散[J].软件学报, 2007, 18(2): 600-607.
[43] Yang C, Zheng S, Ye J. Level set contour extraction method based on support value filter [J]. Applied Mathematics and Computation, 2008, 205: 688-696.
[44]冯玉玲,何传江,李梦.不用高斯平滑的边缘活动轮廓模型[J].计算机工程与应用, 2010, 46 (36): 192-194.
[45]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,2001.29～35.