马尔可夫随机场在连续太赫兹波图像处理中的应用研究
|
摘要
太赫兹(THz)成像以其较低的光子能量和较高的穿透能力,在医学检查、安全检测、环境监测和空间遥感等方面展现出巨大的前景。连续THz波成像系统虽不具备脉冲THz波的光谱成像能力,但是它结构简单,操作方便,并且具有相对较低的成本,快速的数据获取能力,高的信噪比,因此具有广阔的应用前景和推广价值。
随着连续THz波成像研究的不断深入以及各种成像技术实用化的需求,连续THz波图像处理技术正逐渐成为一个新的研究领域。从图像中识别出有用的信息几乎是所有成像系统的最终目的,然而利用成像设备获得的原始图像并不能直观真实地反映目标的信息。噪声等因素的影响使图像偏离了真实情况,极大地影响了人们从图像中提取信息,因此非常有必要在利用图像之前对图像进行处理。本论文以连续THz波图像为主要研究对象,分析了图像的主要特征,并将马尔可夫随机场(MRF)的方法引入到图像处理中,进行了相关的应用研究。
本论文首先从连续THz波图像的噪声、对比度、分辨率和成像条纹等方面对图像进行了分析,并介绍了相关实用的处理算法。利用像素行平均的方法可以抑制由于激光输出功率缓慢波动而引入的图像噪声;以处理后的THz激光光斑作为点扩散函数,利用Lucy-Richardson图像恢复算法可以增强THz图像的分辨率;利用基于离散余弦变换(DCT)的一维滤波方法可以对图像条纹频率进行提取并滤除条纹。
本文通过分析连续THz波扫描图像的邻域相关性,结合MRF的定义,证实了连续THz波扫描图像的马尔可夫性,并建立了THz图像的MRF模型,给出了用于THz图像去噪和分割处理的能量函数。结合随机松驰算法(模拟退火算法)实现了THz图像MRF模型的优化过程。理论模型与算法结合的有效性在连续波THz图像去噪与分割处理中得到了有效的验证。
Terahertz (THz) imaging with the characteristics of lower photons energy and higher penetrability shows an expansive future in medical inspection, safety detection, environmental monitoring and space remote sensing. Though unlike the case of pulsed THz, a continuous-wave (CW) THz imaging system lacks spectral information, with the advantages of simple structure, easy in operation, lower cost, fast data acquisition capacity and higher signal noise rate, it shows wide application foreground and value.
With the development of CW THz imaging research and practical requirement of various imaging techniques, the CW THz image processing technique is growing to be a new research field. To identify useful information in an image is the ultimate objective of an imaging system. But the original images captured by an imaging system can’t reflect the true information, which influences the information acquisition from the original images. So it is very necessary to process the original images before application. In this thesis, focusing on CW THz image, the characteristics of the image have been analyzed and some applied researches have been carried out through introducing Markov random field (MRF) to THz image processing.
In this thesis, the image noise, contrast, resolution and fringes of CW THz image have been firstly analyzed and some useful algorithms have been introduced. The noise caused by the laser output power fluctuation is suppressed with the algorithm of line-pixel average. The resolution of THz images can be enhanced by Lucy-Richardson algorithm through using the THz imaging spot as the point spread function (PSF). The frequency band of the fringes can be extracted and the image fringes can be removed by one dimension filtering based on discrete cosine transform (DCT).
In this thesis, with the analysis of the pixel’s neighborhood correlation of THz image and the definition of MRF, the CW THz scanning image is proved to obey to the definition of MRF. The image’s MRF model is established and the energy functions which are used for image denoising and segmentation are given and optimized by simulated annealing algorithm. The combination effectivity of theory model and algorithm is validated in the application of THz image denoising and segmentation.
随着连续THz波成像研究的不断深入以及各种成像技术实用化的需求,连续THz波图像处理技术正逐渐成为一个新的研究领域。从图像中识别出有用的信息几乎是所有成像系统的最终目的,然而利用成像设备获得的原始图像并不能直观真实地反映目标的信息。噪声等因素的影响使图像偏离了真实情况,极大地影响了人们从图像中提取信息,因此非常有必要在利用图像之前对图像进行处理。本论文以连续THz波图像为主要研究对象,分析了图像的主要特征,并将马尔可夫随机场(MRF)的方法引入到图像处理中,进行了相关的应用研究。
本论文首先从连续THz波图像的噪声、对比度、分辨率和成像条纹等方面对图像进行了分析,并介绍了相关实用的处理算法。利用像素行平均的方法可以抑制由于激光输出功率缓慢波动而引入的图像噪声;以处理后的THz激光光斑作为点扩散函数,利用Lucy-Richardson图像恢复算法可以增强THz图像的分辨率;利用基于离散余弦变换(DCT)的一维滤波方法可以对图像条纹频率进行提取并滤除条纹。
本文通过分析连续THz波扫描图像的邻域相关性,结合MRF的定义,证实了连续THz波扫描图像的马尔可夫性,并建立了THz图像的MRF模型,给出了用于THz图像去噪和分割处理的能量函数。结合随机松驰算法(模拟退火算法)实现了THz图像MRF模型的优化过程。理论模型与算法结合的有效性在连续波THz图像去噪与分割处理中得到了有效的验证。
Terahertz (THz) imaging with the characteristics of lower photons energy and higher penetrability shows an expansive future in medical inspection, safety detection, environmental monitoring and space remote sensing. Though unlike the case of pulsed THz, a continuous-wave (CW) THz imaging system lacks spectral information, with the advantages of simple structure, easy in operation, lower cost, fast data acquisition capacity and higher signal noise rate, it shows wide application foreground and value.
With the development of CW THz imaging research and practical requirement of various imaging techniques, the CW THz image processing technique is growing to be a new research field. To identify useful information in an image is the ultimate objective of an imaging system. But the original images captured by an imaging system can’t reflect the true information, which influences the information acquisition from the original images. So it is very necessary to process the original images before application. In this thesis, focusing on CW THz image, the characteristics of the image have been analyzed and some applied researches have been carried out through introducing Markov random field (MRF) to THz image processing.
In this thesis, the image noise, contrast, resolution and fringes of CW THz image have been firstly analyzed and some useful algorithms have been introduced. The noise caused by the laser output power fluctuation is suppressed with the algorithm of line-pixel average. The resolution of THz images can be enhanced by Lucy-Richardson algorithm through using the THz imaging spot as the point spread function (PSF). The frequency band of the fringes can be extracted and the image fringes can be removed by one dimension filtering based on discrete cosine transform (DCT).
In this thesis, with the analysis of the pixel’s neighborhood correlation of THz image and the definition of MRF, the CW THz scanning image is proved to obey to the definition of MRF. The image’s MRF model is established and the energy functions which are used for image denoising and segmentation are given and optimized by simulated annealing algorithm. The combination effectivity of theory model and algorithm is validated in the application of THz image denoising and segmentation.
引文
1 N. I. Malykh, A. G. Nagornyi, and E. S. Yampolskii. Submillimeter-wave imagery. Instruments and Experimental Techniques, 1975, 18(1): 182-184
2 T. S. Hartwick, D. T. Hodges, D. H. Barker et al. Far Infrared Imagery. Applied Optics, 1976, 15(8): 1919-1922
3 H. M. pickett, R. L. Poynter, E. A. Cohen. Submillimeter, Millimeter and microwave spectral line catalog, http://spec.jpl.Nasa.gov(2003)
4 H. M. pickett, R. L. Poynter, E. A. Cohen et al. Submillimeter. Millimeter and Microwave spectral line catalog. Journal of Quantitative Spectroscopy and Radiation Transfer, 1998, (60): 883-890
5 C. M. Mann, P. D. Maagt, G. Mcbride et al. Microfabrication of 3D Terahertz Circuitry. IEEE MIT-S International Microwave Symposium Digest, Piscataway, NJ, 2003, 2: 739-742
6 John F. Federici, Dale Gary, Robert Barat, David Zimdars. THz Standoff Detection and Imaging of Explosives and Weapons. Proc. SPIE 5781, 2005, (75)
7 THz imaging of space shuttle foam, Nature, August, 2003, (14)
8谢旭,钟化,袁韬等.使用太赫兹技术研究航天飞机失事的原因.物理, 2003, 32(9): 583-584
9迟欣.太赫兹激光透射成像技术初步研究.哈尔滨工业大学硕士学位论文, 2007: 38-59
10单纪鑫.太赫兹扫描反射成像与实时透射成像初步研究.哈尔滨工业大学硕士学位论文, 2008: 16-36
11 Caihong Zhang, YuanYuan Wang, Jian Chen et al. Continous-wave Terahertz Imaging System Based on Far-infrared Laser Source. Conference Digest of the 2006 Joint 31st International Conference on Infrared and Millimeter Waves and 14th International Conference on Terahertz Electronics, 2006:426
12杨珺,阮双琛,张敏,张伟.实时连续波太赫兹成像在邮件检测中的应用.深圳大学学报(理工版), 2008, 25(3): 261-263
13张敏,阮双琛,杨珺等.连续太赫兹波实时透射成像实验研究.深圳大学学报(理工版), 2007, 24(4): 384-387
14周燕,牧凯军,张艳东等.燃料箱泡沫板的连续太赫兹波无损检测.无损检测, 2007, 25(5): 266-268
15张艳东.连续太赫兹波成像技术的检测应用研究.首都师范大学硕士学位论文, 2008: 16-20
16周燕.连续太赫兹波成像技术的检测应用研究.首都师范大学硕士学位论文, 2007
17焦月英.连续太赫兹波成像技术的研究.首都师范大学硕士学位论文, 2008
18 Kave Houshmand, Hamid Reza Tizhoosh. Filtering and Fusion of THz Images for Defect Detection in Composite Materials. 2008 IEEE International Conference on Fuzzy Systems, 2008, 1301-1305
19 Yu Fei, Hui Mei, Song Qian, Zhao Yue-jin. The Application of the Wavelet Transform to the Continuous Wave Terahertz Imaging. Optical Fiber Communication and Optoelectronics Conference, Asia, 2007, 502-504
20李海涛.连续太赫兹波图像增强算法的研究.首都师范大学硕士学位论文. 2008
21李琦,姚睿,单纪鑫,尹奇国,王骐.基于热像仪的太赫兹成像及图像增强初步研究[J].光学学报, 2008, 28(s2): 275～277
22 K Abend, T J Harley, L N Kanal. Classification of Binary Random Patterns. IEEE Trans Inform Theory IT-11, 1965: 538-544
23 N Ahuja. Mosaic Models for Image Analysis and Synthesis. Ph D Dissertaton. Department of Computer Science, University of Maryland, College Park, Maryland, 1979
24 Laveen N Kanal. Markov Mesh models. Image Modeling, New York: Academic Press, 1980: 240-243
25 S Geman, D Geman. Stochastic Relaxation Gibbs Distribution and the Bayesian Restoration of Images. IEEE Trans Pattern Anal Machine Intell, 1984, PAMI-16: 721-741
26 D Geiger, F Girosi. Parallel and Deteministic Algorithms from MRF’s: Surface Reconstruction. IEEE Trans PAMI, 1991, 13(5): 401-412
27 H Derin, H Elliott. Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields. IEEE Trans PAMI, 1987, 9(1): 39-55
28 B S Manjunath, R Chellappa. Unsupervised Texture Segmentation Using Markov Random Field Models. IEEE Trans PAMI, 1991, 13(5): 478-482
29 M Daily. Color Image Segmentation Using Markov Random Field. Proceedings IEEE Conference of Computer Vision and Pattern Recognition, 1989: 304-312
30 V E Johnson, W H Wong, X Hu et al. Image Restoration Using Gibbs Priors: Boundary Modeling, Treatment of Blurring, and Selection of Hyper-parameter. IEEE Trans PAMI, 1991, 13(5): 413-425
31 D Geman, S Geman, C Graffigne et al. Boundary Detection by Constrained Optimization. IEEE Trans PAMI, 1990, 12(7): 609-628
32 R C Dubles, A K Jain, S G Nadabar et al. MRF Model-based Algorithms for Image Segmentation. Proceedings IEEE 10th International Conference on Pattern Recognition, Atlantic City, NJ,1990(06): 808-814
33 A K Jain, S G Nadabar. MRF Model-based Segmentation of Range Images. Proceedings International Conference on Computer Vision, 1990: 667-671
34 S G Nadabar, A K Jain. Parameter Estimation in MRF Line Process Models. Proceedings IEEE Conference on Computer Vision and Pattern Recognition, 1992(06): 528-533
35 J E Besag. Spatial Interaction and the Statistical Analysis of Lattice System. J R Stat Soc B-36, 1974: 192-236
36 Yongyue Zhang, Michael Brady, Stephen Smith. Segmentation of Brain MR Images Through a Hidden Markov Random Field Model and the Expectation-Maximization Algorithm. IEEE Transations on Medical Imaging. 2001, 20(1): 45-57
37 Max Mignotte, Christophe Collet, Patrick Perez, Patrick Bouthemy, Sonar Image Segmentation Using an Unsupervised Hierarchical MRF Model. IEEE Transactions on Image Processing, 2000, 9(7): 1216-1232
38杨子宁.基于马尔可夫随机场的降质光学图像预处理.电光与控制, 2007, 14(6): 183-185
39侯一民,郭雷.一种基于马尔可夫随机场的SAR图像分割新方法.电子与信息学报, 2007, 29(5): 1069-1072
40 Huawu Deng, David A. Clausi, Unsupervised Image Segmentation Using a Simple MRF Model with a New Implementation Scheme. Pattern Recognition, 2004, (37): 2323-2335
41 Peter H. Handel, Amanda M. Hall, Adam G. Tournier. Noise in THz Detectors and Generators. Proc. of SPIE, 2006, 6212: 62120D1-62120D12
42 Huang Lin, Tao Cunkan, Hu Maohai. A Blind Deconvolution Method in a LCSM System. Proc. of SPIE, 2006, 6149: 614932D1-614932D5
43 T. Hattori, M. Skakmoto, R. Rungsawang. Spatial Frequency Filter for Real-time THz Imaging. Infrared and Millimeter Waves and 13th International Conference on Terahertz Electronics, 2005, 2: 333-334
44 Nick Karpowicz. Removal of Standing Wave Patters from CW THz Images through Spatial Frequency Filtering. Project Report of Standing Wave Filtering, 15th May 2006: 1-8
45郑肇葆.图像分析的马尔可夫随机场方法.武汉测绘科技大学出版社,2000
46 Spitzer. Markov Random Field and Gibbs Ensembles. Amer Math Mon, 1971(78): 152-154
47王平,牟轩心,王勇,俞恒永.基于马尔可夫随机场的图像恢复方法.计算机工程与应用, 2001, (13): 100-101
48李华北,赵杰文.基于Markov随机场的猪肉糜微结构图像模拟.农业机械学报, 2001, 32(2): 62-65
49 Scott T. Sction. Image Restoration Using Generalized Deterministic Annealing. Digital Signal Processing, 1997, (7): 94-104
50 Christine Haas. Segmentation of 3D Intravascular Ultrasonic Images Based on a Random Field Model. Ultrasound in Med. & Biol., 2000, 26(2): 297-306
51 Maurits Malfait, Dirk Roose. Wavelet-Based Image Denoising Using a Markov Random Field a Priori Model. IEEE Transactions on Image Processing, 1997, 6(4)
52刘忠,茆诗松.分组数据的Bayes分析—Gibbs抽样方法.应用概率统计, 1997, 13(2): 211-216
53龙翔. Bayesian运图像分析.西北工业大学博士学位论文, 2003
54 Georgy Gimel’farb. Modeling Images Textures by Gibbs Random Fields. Pattern Recognition Letters, 1999, (20): 1123-1132
55夏正良.数字图像处理.东南大学出版社, 1999: 60-61
56 M. Miqnotte, C. Collet, P.Perez, P. Bouthemy. Unsupervised Markovian Segmentation of Sonar Images. Acoustics, Speech, and Signal Processing, IEEE, 1997,4: 2781-2784
57 Tamas Sziranyi, Josiane Zerubia. Markov Random Field Image Segmentation Using Cellular Neural Network. IEEE Transactions on Circuits and Systems: Fundamental Theory and Applications, 1997, 44(1): 86-89
2 T. S. Hartwick, D. T. Hodges, D. H. Barker et al. Far Infrared Imagery. Applied Optics, 1976, 15(8): 1919-1922
3 H. M. pickett, R. L. Poynter, E. A. Cohen. Submillimeter, Millimeter and microwave spectral line catalog, http://spec.jpl.Nasa.gov(2003)
4 H. M. pickett, R. L. Poynter, E. A. Cohen et al. Submillimeter. Millimeter and Microwave spectral line catalog. Journal of Quantitative Spectroscopy and Radiation Transfer, 1998, (60): 883-890
5 C. M. Mann, P. D. Maagt, G. Mcbride et al. Microfabrication of 3D Terahertz Circuitry. IEEE MIT-S International Microwave Symposium Digest, Piscataway, NJ, 2003, 2: 739-742
6 John F. Federici, Dale Gary, Robert Barat, David Zimdars. THz Standoff Detection and Imaging of Explosives and Weapons. Proc. SPIE 5781, 2005, (75)
7 THz imaging of space shuttle foam, Nature, August, 2003, (14)
8谢旭,钟化,袁韬等.使用太赫兹技术研究航天飞机失事的原因.物理, 2003, 32(9): 583-584
9迟欣.太赫兹激光透射成像技术初步研究.哈尔滨工业大学硕士学位论文, 2007: 38-59
10单纪鑫.太赫兹扫描反射成像与实时透射成像初步研究.哈尔滨工业大学硕士学位论文, 2008: 16-36
11 Caihong Zhang, YuanYuan Wang, Jian Chen et al. Continous-wave Terahertz Imaging System Based on Far-infrared Laser Source. Conference Digest of the 2006 Joint 31st International Conference on Infrared and Millimeter Waves and 14th International Conference on Terahertz Electronics, 2006:426
12杨珺,阮双琛,张敏,张伟.实时连续波太赫兹成像在邮件检测中的应用.深圳大学学报(理工版), 2008, 25(3): 261-263
13张敏,阮双琛,杨珺等.连续太赫兹波实时透射成像实验研究.深圳大学学报(理工版), 2007, 24(4): 384-387
14周燕,牧凯军,张艳东等.燃料箱泡沫板的连续太赫兹波无损检测.无损检测, 2007, 25(5): 266-268
15张艳东.连续太赫兹波成像技术的检测应用研究.首都师范大学硕士学位论文, 2008: 16-20
16周燕.连续太赫兹波成像技术的检测应用研究.首都师范大学硕士学位论文, 2007
17焦月英.连续太赫兹波成像技术的研究.首都师范大学硕士学位论文, 2008
18 Kave Houshmand, Hamid Reza Tizhoosh. Filtering and Fusion of THz Images for Defect Detection in Composite Materials. 2008 IEEE International Conference on Fuzzy Systems, 2008, 1301-1305
19 Yu Fei, Hui Mei, Song Qian, Zhao Yue-jin. The Application of the Wavelet Transform to the Continuous Wave Terahertz Imaging. Optical Fiber Communication and Optoelectronics Conference, Asia, 2007, 502-504
20李海涛.连续太赫兹波图像增强算法的研究.首都师范大学硕士学位论文. 2008
21李琦,姚睿,单纪鑫,尹奇国,王骐.基于热像仪的太赫兹成像及图像增强初步研究[J].光学学报, 2008, 28(s2): 275～277
22 K Abend, T J Harley, L N Kanal. Classification of Binary Random Patterns. IEEE Trans Inform Theory IT-11, 1965: 538-544
23 N Ahuja. Mosaic Models for Image Analysis and Synthesis. Ph D Dissertaton. Department of Computer Science, University of Maryland, College Park, Maryland, 1979
24 Laveen N Kanal. Markov Mesh models. Image Modeling, New York: Academic Press, 1980: 240-243
25 S Geman, D Geman. Stochastic Relaxation Gibbs Distribution and the Bayesian Restoration of Images. IEEE Trans Pattern Anal Machine Intell, 1984, PAMI-16: 721-741
26 D Geiger, F Girosi. Parallel and Deteministic Algorithms from MRF’s: Surface Reconstruction. IEEE Trans PAMI, 1991, 13(5): 401-412
27 H Derin, H Elliott. Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields. IEEE Trans PAMI, 1987, 9(1): 39-55
28 B S Manjunath, R Chellappa. Unsupervised Texture Segmentation Using Markov Random Field Models. IEEE Trans PAMI, 1991, 13(5): 478-482
29 M Daily. Color Image Segmentation Using Markov Random Field. Proceedings IEEE Conference of Computer Vision and Pattern Recognition, 1989: 304-312
30 V E Johnson, W H Wong, X Hu et al. Image Restoration Using Gibbs Priors: Boundary Modeling, Treatment of Blurring, and Selection of Hyper-parameter. IEEE Trans PAMI, 1991, 13(5): 413-425
31 D Geman, S Geman, C Graffigne et al. Boundary Detection by Constrained Optimization. IEEE Trans PAMI, 1990, 12(7): 609-628
32 R C Dubles, A K Jain, S G Nadabar et al. MRF Model-based Algorithms for Image Segmentation. Proceedings IEEE 10th International Conference on Pattern Recognition, Atlantic City, NJ,1990(06): 808-814
33 A K Jain, S G Nadabar. MRF Model-based Segmentation of Range Images. Proceedings International Conference on Computer Vision, 1990: 667-671
34 S G Nadabar, A K Jain. Parameter Estimation in MRF Line Process Models. Proceedings IEEE Conference on Computer Vision and Pattern Recognition, 1992(06): 528-533
35 J E Besag. Spatial Interaction and the Statistical Analysis of Lattice System. J R Stat Soc B-36, 1974: 192-236
36 Yongyue Zhang, Michael Brady, Stephen Smith. Segmentation of Brain MR Images Through a Hidden Markov Random Field Model and the Expectation-Maximization Algorithm. IEEE Transations on Medical Imaging. 2001, 20(1): 45-57
37 Max Mignotte, Christophe Collet, Patrick Perez, Patrick Bouthemy, Sonar Image Segmentation Using an Unsupervised Hierarchical MRF Model. IEEE Transactions on Image Processing, 2000, 9(7): 1216-1232
38杨子宁.基于马尔可夫随机场的降质光学图像预处理.电光与控制, 2007, 14(6): 183-185
39侯一民,郭雷.一种基于马尔可夫随机场的SAR图像分割新方法.电子与信息学报, 2007, 29(5): 1069-1072
40 Huawu Deng, David A. Clausi, Unsupervised Image Segmentation Using a Simple MRF Model with a New Implementation Scheme. Pattern Recognition, 2004, (37): 2323-2335
41 Peter H. Handel, Amanda M. Hall, Adam G. Tournier. Noise in THz Detectors and Generators. Proc. of SPIE, 2006, 6212: 62120D1-62120D12
42 Huang Lin, Tao Cunkan, Hu Maohai. A Blind Deconvolution Method in a LCSM System. Proc. of SPIE, 2006, 6149: 614932D1-614932D5
43 T. Hattori, M. Skakmoto, R. Rungsawang. Spatial Frequency Filter for Real-time THz Imaging. Infrared and Millimeter Waves and 13th International Conference on Terahertz Electronics, 2005, 2: 333-334
44 Nick Karpowicz. Removal of Standing Wave Patters from CW THz Images through Spatial Frequency Filtering. Project Report of Standing Wave Filtering, 15th May 2006: 1-8
45郑肇葆.图像分析的马尔可夫随机场方法.武汉测绘科技大学出版社,2000
46 Spitzer. Markov Random Field and Gibbs Ensembles. Amer Math Mon, 1971(78): 152-154
47王平,牟轩心,王勇,俞恒永.基于马尔可夫随机场的图像恢复方法.计算机工程与应用, 2001, (13): 100-101
48李华北,赵杰文.基于Markov随机场的猪肉糜微结构图像模拟.农业机械学报, 2001, 32(2): 62-65
49 Scott T. Sction. Image Restoration Using Generalized Deterministic Annealing. Digital Signal Processing, 1997, (7): 94-104
50 Christine Haas. Segmentation of 3D Intravascular Ultrasonic Images Based on a Random Field Model. Ultrasound in Med. & Biol., 2000, 26(2): 297-306
51 Maurits Malfait, Dirk Roose. Wavelet-Based Image Denoising Using a Markov Random Field a Priori Model. IEEE Transactions on Image Processing, 1997, 6(4)
52刘忠,茆诗松.分组数据的Bayes分析—Gibbs抽样方法.应用概率统计, 1997, 13(2): 211-216
53龙翔. Bayesian运图像分析.西北工业大学博士学位论文, 2003
54 Georgy Gimel’farb. Modeling Images Textures by Gibbs Random Fields. Pattern Recognition Letters, 1999, (20): 1123-1132
55夏正良.数字图像处理.东南大学出版社, 1999: 60-61
56 M. Miqnotte, C. Collet, P.Perez, P. Bouthemy. Unsupervised Markovian Segmentation of Sonar Images. Acoustics, Speech, and Signal Processing, IEEE, 1997,4: 2781-2784
57 Tamas Sziranyi, Josiane Zerubia. Markov Random Field Image Segmentation Using Cellular Neural Network. IEEE Transactions on Circuits and Systems: Fundamental Theory and Applications, 1997, 44(1): 86-89