工业射线图像增强算法研究
|
摘要
当今社会中,X射线检测在安检、医学、工业探伤等方面发挥着不可或缺的作用,已成为推动国民经济发展的一支重要力量。然而,由于X射线检测系统固有的缺陷,输出后的图像往往要经过处理才能达到要求。本文结合项目组的新型X射线成像系统,对工业成像的增强算法做了一些研究,提出了一些改进的图像增强算法,主要工作如下:
1.比较详细地介绍了常规的图像增强算法,一般可分为两大类,即空域和频域增强,针对射线图像的特点,有选择地采用了不同增强算法,进行了实验验证,结果表明,常规增强算法对射线图像是有一定增强效果的;
2.系统地介绍了模糊增强算法,它已成为近年来研究的热点,在传统模糊增强的基础上,本文给出了五种改进的单层次模糊增强算法,然后介绍了多层次模糊增强算法,而且每种算法都和原算法进行了定量比较,实验结果表明,改进后的算法对图像的增强效果要优于原算法;
3.设计了基于小波域的模糊增强算法,它是本文的重点,本文使用了两种不同的算法来实现小波域上的模糊增强,第一种是直接对小波分解后的低频图像采用模糊增强算法,第二种是对小波增强后的图像使用模糊增强算法,详细给出了每种算法的实现步骤,而且定量比较了两种算法的优劣,增强后的图像表明,使用此类算法要优于单纯使用模糊增强和小波增强算法;
4.把反锐化掩模和形态学边缘检测结合起来,构建了一个新的算法,它弥补了各自算法的不足,而且在原算法的基础上,本文都做了进一步改进和提高,采用此算法和传统的边缘检测算法增强图像作了一个比较,实验表明,此算法可以显示更多细节,边缘检测效果更明显,达到了预期目的。
Nowadays,X-ray image plays an important role in the field of secure detection,mediation and industry detection etc.It becomes an important impluse for the development of country.Because of the X-ray system inherent defect, the X-ray image must be handled in the computer.Based on the new X-ray image system,we mainly study X-ray image enhancement algorithms and gives some improved algorithms.The paper mainly include :
1. It introduces the usual image enhancement theory.This kind of algorithm can be divided into two categories:airspace and frequency enhancement .Based on the X-ray image,we selectively does some experiments. These algorithms have some effects the on X-ray image enhancement.
2. It systemly introduces the traditonal fuzzy enhancement and gives four improved algorithms.Else,it gives improved generalized fuzzy enhancement algorithm and multi-level fuzzy enhancement algorithm.For each algorithm,it gives quantitative comparation between the former algorithm and the improved algorithm.The experiment shows the improved algorithm better than the former algorithm.
3. It creates a kind of algorithm that is how to use fuzzy sets to realize image enhancement based on wavelet transform.The first algorithm uses fuzzy enhancement for low frequency of wavelet transform imageThe second algorithm uses fuzzy enhancement for image of wavelet transform enhancment.It gives a detailed description for each algorithm.What’s more,it gives quantitative comparation between two algorithms. The experiment shows that it can produce a best result of all.
4. It uses unsharpening mask and mathematical morphology edge detection to design a new algorithm to realize image enhancement.It compensates defects of each algorithm.Moreover,it improves each former algorithm .Compared with the traditional edge enhancement , this algorithm makes the X-ray image more clear. The result shows that this method is feasible and useful for X-ray image enhancement.
1.比较详细地介绍了常规的图像增强算法,一般可分为两大类,即空域和频域增强,针对射线图像的特点,有选择地采用了不同增强算法,进行了实验验证,结果表明,常规增强算法对射线图像是有一定增强效果的;
2.系统地介绍了模糊增强算法,它已成为近年来研究的热点,在传统模糊增强的基础上,本文给出了五种改进的单层次模糊增强算法,然后介绍了多层次模糊增强算法,而且每种算法都和原算法进行了定量比较,实验结果表明,改进后的算法对图像的增强效果要优于原算法;
3.设计了基于小波域的模糊增强算法,它是本文的重点,本文使用了两种不同的算法来实现小波域上的模糊增强,第一种是直接对小波分解后的低频图像采用模糊增强算法,第二种是对小波增强后的图像使用模糊增强算法,详细给出了每种算法的实现步骤,而且定量比较了两种算法的优劣,增强后的图像表明,使用此类算法要优于单纯使用模糊增强和小波增强算法;
4.把反锐化掩模和形态学边缘检测结合起来,构建了一个新的算法,它弥补了各自算法的不足,而且在原算法的基础上,本文都做了进一步改进和提高,采用此算法和传统的边缘检测算法增强图像作了一个比较,实验表明,此算法可以显示更多细节,边缘检测效果更明显,达到了预期目的。
Nowadays,X-ray image plays an important role in the field of secure detection,mediation and industry detection etc.It becomes an important impluse for the development of country.Because of the X-ray system inherent defect, the X-ray image must be handled in the computer.Based on the new X-ray image system,we mainly study X-ray image enhancement algorithms and gives some improved algorithms.The paper mainly include :
1. It introduces the usual image enhancement theory.This kind of algorithm can be divided into two categories:airspace and frequency enhancement .Based on the X-ray image,we selectively does some experiments. These algorithms have some effects the on X-ray image enhancement.
2. It systemly introduces the traditonal fuzzy enhancement and gives four improved algorithms.Else,it gives improved generalized fuzzy enhancement algorithm and multi-level fuzzy enhancement algorithm.For each algorithm,it gives quantitative comparation between the former algorithm and the improved algorithm.The experiment shows the improved algorithm better than the former algorithm.
3. It creates a kind of algorithm that is how to use fuzzy sets to realize image enhancement based on wavelet transform.The first algorithm uses fuzzy enhancement for low frequency of wavelet transform imageThe second algorithm uses fuzzy enhancement for image of wavelet transform enhancment.It gives a detailed description for each algorithm.What’s more,it gives quantitative comparation between two algorithms. The experiment shows that it can produce a best result of all.
4. It uses unsharpening mask and mathematical morphology edge detection to design a new algorithm to realize image enhancement.It compensates defects of each algorithm.Moreover,it improves each former algorithm .Compared with the traditional edge enhancement , this algorithm makes the X-ray image more clear. The result shows that this method is feasible and useful for X-ray image enhancement.
引文
1.藤田渑等.日本造船科技报.39(4):1-6.
2.G.Robisen,Bennett.An Enhancement Digited Video Images for Improved NDT[J]. Mental Construction. 1985, 17(11):15-18.
3.X-Ray Real-time Imaging for Weld Inspection.W2nd Progress Report British Journal of NDT[J].London:1987,29(1):22-23.
4.X-Ray Real-time Imaging for Weld Inspection.W3nd Summary report British Journal of NDT. London:1990, 32(1):9-13.
5.Just T,Thale W Clausen et al. Interpretation of radiographs by didital image processing[J/OL].The E-Journal of Nondestructive Testing.1998,11(3):48-52.
6.Dau H D.Digital image Processing in X-ray control of defect[J/OL]. The E-Journal of Nondestructive Testing.1998,26(5):75-79.
7.周伟,余华民.图像处理和模式识别在射线检验中的应用[J].无损探伤.1990, 14(5):1-5.
8.孙忠诚等.X射线实时探伤数字图像处理研究[J].无损探伤.1992 ,14(2):37-39.
9.钢铁,王东华.X射线检测图像中缺陷的自动提取和分割[J].焊接学报.2001,13(5): 6-9.
10.孙怡,孙洪雨,白鹏等.射线焊缝图像实时检测方法[J].焊接学报.2004 ,25(2):41-44.
11.Stark J.A.Adpative image contrast enhancement using generalizations of histogram equalization.IEEE Trans Image Processing. 2000,9(5):889-896.
12.蔡超.基于小波和偏微分方程的图像处理方法与应用.博士论文.武汉:华中科技大学,2005.
13.喻春雨.新型X光机影像探测器及成像系统研究.博士论文.南京:南京理工大学,2006.
14.陈树越,路宏年.X射线数字成像噪声特性及噪声消除方法研究[J].无损检测.2001,14(1):47-50.
15.焦建兵,张军杰,常本康,富容国.数字化X光机系统研究.中国兵工学会第四届夜市技术学术交流会论文集.北京:2003:95-99.
16.冯卓煌.射线图像处理算法研究与硬件实现.硕士论文.南京:南京理工大学,2007.
17.姚敏.数字图象处理[M].北京:机械工业出版社.2007:36-37.
18.赵书兰主编.Matlab R2008数字图像处理与分析实例教程[M].北京:化学工业出版社.2009:72-74.
19.谢勤岚.图像降噪的自适应高斯平滑滤波器[J].计算机工程与应用.2009, 45(6) :182-184.
20.PrattWR.Digital image processing [M].NewYork.1977:201-203.
21.刘劲,康志伟,何怡刚.基于小波变换的反锐化掩模图像增强新算法[J].计算机工程.2007, 33(21):183-185.
22.S K PAL,R A King. Image enhancement using smoothing with fuzzy set[C].IEEE Trans Syst Man Cybern,1981,SMC-11(7):494-501.
23.http://watfor.uwaterloo.ca/tizhoosh/history.htm.
24.Osama K .Alshaykh,Srikrishna Ramaswamy Hesen-Sen Hung.Fuzzy Techniques forimage enhancement and Reconstruction.Second IEEE International Conference on Fuzzy Systems.1993,6(1):582-587.
25.Sankar K .Pal.Fuzzy sets in image processing and recognition.IEEE International Conference on fuzzy systems.1992,10(2):119-126.
26.董远,刘文萍,孙肖子.模糊理论在图象增强中的应用[J].数据采集与处理.1995,7(20) :25-28.
27.刘兴伟,殷国富,李双跃.自适应的模糊增强算法及其在工程扫描图中的应用[J].机械设计.2001,1(28) :69-72.
28.饶娟.基于模糊增强图象边缘检测的研究[J].甘肃联合大学学报.2006,6(27): 46-51.
29.雷向康.一种改进的图像模糊增强算法[J].系统工程与电子技术1997,12(3):21-25.
30.刘习文,蒋艳荣,罗显光.一种改进的图像模糊增强算法[J].计算机工程与应用.2008,44(4): 50-52.
31.高顶,刘元祥,张晓光.基于广义模糊算子的工业RT图像增强[J].华东理工大学学报.2006 ,32(3):332-334.
32.石振刚,高立群.一种基于模糊逻辑的图像多层次增强算法[J].控制工程.2006 ,13(5) :463-465.
33. Menotti D,Najiman L,Facon J et al. Multi-histogram equalization methods for contrast enhancement and brightness preserving. IEEE Transactions on Consumer Electronics.2007, 53 (3):1186-1194.
34.Grundland M,Dodgson N A. Interactive contrast enhancement by histogram warping[C] .International Conference on Computer Vision and Graphics. 2004,32 :832-838.
35.王保平,刘生虎,范九伦.基于模糊熵的自适应图像多层次模糊增强算法[J].电子学报.2005 ,33(4):730-734.
36.刘博,胡正平,王成儒.基于模糊松弛迭代的分层图像增强算法[J].光学技术.2009,35(1):134-137.
37.Morlet.Wavelet Transform for earthquake analysis[J].Signal Processing.1980, 7(2):1055-1061.
38.Arenodo,G.Grasseau.Wavelet Transform of Multifractals[J].Physical Review letters.1988, 6(1):2281-2284.
39.Jeon,Y.S.Shin.Wavelet Transform for time-frequency Analysis of the Vibration Signature and its Application[J] . IEEE Trans Image Processing. 1996,6(4):1302-1306.
40.Rose. A new algorithm about wavelet hard thresholding for X-ray image defection . IEEE Trans Image Processing .2006, 15(9):13-19.
41.Puillion.Wavelet Transform for X-ray image defection[J].IEEE Trans Image Processing.2006, 23(1):111-117.
42.钢铁.基于神经网络的智能化超声焊接缺陷模式识别与诊断.硕士论文.哈尔滨:哈尔滨工业大学,1995.
43.张广明.超声无损检测中的时频分析理论及应用研究.西安:西安交通大学1999,91-94.
44.刘镇清.超声无损检测与评价中的信号处理及模式识别[J].无损检测2001,23(6):31-34.
45.王恩亮.X射线无损探伤中的缺陷的自动检测和识别.大连:大连理工大学, 2003:73-77.
46.Kenneth R.Castleman著,朱志刚等译.数字图象处理[M].北京:电子工业出版社,2006:265-267.
47.[法]Stephane Mallat著.杨力华,戴道清,黄文良等译.信号处理的小波导引[M].北京:机械工业出版社,2002:106-109.
48.王彬.熵与信息[M].西安:西北工业大学出版社,1994:154-155.
49.周延武,李保国,宗光华.小波模糊理论在擦窗机器人实时图象检测中的应用[J].机器人.2002,24[5]: 26-30.
50.徐碧.基于小波变换的图象增强方法研究.硕士论文.长春:中国科学院光电技术研究所,1999.
51.薛琴,邹道文,吕珂.基于模平方处理的小波图像增强方法[J].江西师范大学学报.2005, 37(6): 276-279.
52.Bovis,Singh.A algorithm about distribution separation measurement for contrast enhancement[J].IEEE Trans Image Processing.1994, 7(4):253-259.
53.Margagos P.Differential and image processing[C].IEEE Trans Image Processing.1996 ,5(6):922-937.
54.杨述斌.数学形态学在数字图像处理中的应用研究.博士论文.武汉:华中科技大学,2002.
55.Canny.A Computational approach to edge detection[J].IEEE Transon PAMI.1985 ,8(6):679-698.
56.Se`bastien Guillion et al.Adaptive nonlinear filters 2D and 3D image enhancement[J].Signal Processing.1998,6(7):237-240.
57.DericheR.Using Canny’s criteria to derive are cursively implemented optimal edge detection[J].The International Journal of Computer Vision.1987, 12(1):167-187.
58.杨述斌,彭复员.噪声污染图像中的广义形态边缘检测器[J].计算机工程与应用.2002,38(7):91-92.
59.杨述斌,彭复员,张增常.多尺度自适应加权形态边缘检测器[J].华中科技大学学报.2002,30(10):42-45.
60.胡晓辉,吕登峰,李敬文.一种基于形态学多结构元的自适应边缘检测算法[J].计算机应用研究.2009,26(2): 752-755.
61.奕新,朱铁一.二次滤波法提取图像边缘信息方法及应用[J].青岛海洋大学学报.1999,29(1):107-111.
62.苑玮琪,王建军,张宏勋.一种基于梯度极值的边缘检测算法[J].信息与控制.1997,26(2):117-120.
63.Stephane Mallat,WenLiang Huang.Singularity detection and processing with wavelets[J].IEEE Transon Information Theory.1992,38(2):617-643.
2.G.Robisen,Bennett.An Enhancement Digited Video Images for Improved NDT[J]. Mental Construction. 1985, 17(11):15-18.
3.X-Ray Real-time Imaging for Weld Inspection.W2nd Progress Report British Journal of NDT[J].London:1987,29(1):22-23.
4.X-Ray Real-time Imaging for Weld Inspection.W3nd Summary report British Journal of NDT. London:1990, 32(1):9-13.
5.Just T,Thale W Clausen et al. Interpretation of radiographs by didital image processing[J/OL].The E-Journal of Nondestructive Testing.1998,11(3):48-52.
6.Dau H D.Digital image Processing in X-ray control of defect[J/OL]. The E-Journal of Nondestructive Testing.1998,26(5):75-79.
7.周伟,余华民.图像处理和模式识别在射线检验中的应用[J].无损探伤.1990, 14(5):1-5.
8.孙忠诚等.X射线实时探伤数字图像处理研究[J].无损探伤.1992 ,14(2):37-39.
9.钢铁,王东华.X射线检测图像中缺陷的自动提取和分割[J].焊接学报.2001,13(5): 6-9.
10.孙怡,孙洪雨,白鹏等.射线焊缝图像实时检测方法[J].焊接学报.2004 ,25(2):41-44.
11.Stark J.A.Adpative image contrast enhancement using generalizations of histogram equalization.IEEE Trans Image Processing. 2000,9(5):889-896.
12.蔡超.基于小波和偏微分方程的图像处理方法与应用.博士论文.武汉:华中科技大学,2005.
13.喻春雨.新型X光机影像探测器及成像系统研究.博士论文.南京:南京理工大学,2006.
14.陈树越,路宏年.X射线数字成像噪声特性及噪声消除方法研究[J].无损检测.2001,14(1):47-50.
15.焦建兵,张军杰,常本康,富容国.数字化X光机系统研究.中国兵工学会第四届夜市技术学术交流会论文集.北京:2003:95-99.
16.冯卓煌.射线图像处理算法研究与硬件实现.硕士论文.南京:南京理工大学,2007.
17.姚敏.数字图象处理[M].北京:机械工业出版社.2007:36-37.
18.赵书兰主编.Matlab R2008数字图像处理与分析实例教程[M].北京:化学工业出版社.2009:72-74.
19.谢勤岚.图像降噪的自适应高斯平滑滤波器[J].计算机工程与应用.2009, 45(6) :182-184.
20.PrattWR.Digital image processing [M].NewYork.1977:201-203.
21.刘劲,康志伟,何怡刚.基于小波变换的反锐化掩模图像增强新算法[J].计算机工程.2007, 33(21):183-185.
22.S K PAL,R A King. Image enhancement using smoothing with fuzzy set[C].IEEE Trans Syst Man Cybern,1981,SMC-11(7):494-501.
23.http://watfor.uwaterloo.ca/tizhoosh/history.htm.
24.Osama K .Alshaykh,Srikrishna Ramaswamy Hesen-Sen Hung.Fuzzy Techniques forimage enhancement and Reconstruction.Second IEEE International Conference on Fuzzy Systems.1993,6(1):582-587.
25.Sankar K .Pal.Fuzzy sets in image processing and recognition.IEEE International Conference on fuzzy systems.1992,10(2):119-126.
26.董远,刘文萍,孙肖子.模糊理论在图象增强中的应用[J].数据采集与处理.1995,7(20) :25-28.
27.刘兴伟,殷国富,李双跃.自适应的模糊增强算法及其在工程扫描图中的应用[J].机械设计.2001,1(28) :69-72.
28.饶娟.基于模糊增强图象边缘检测的研究[J].甘肃联合大学学报.2006,6(27): 46-51.
29.雷向康.一种改进的图像模糊增强算法[J].系统工程与电子技术1997,12(3):21-25.
30.刘习文,蒋艳荣,罗显光.一种改进的图像模糊增强算法[J].计算机工程与应用.2008,44(4): 50-52.
31.高顶,刘元祥,张晓光.基于广义模糊算子的工业RT图像增强[J].华东理工大学学报.2006 ,32(3):332-334.
32.石振刚,高立群.一种基于模糊逻辑的图像多层次增强算法[J].控制工程.2006 ,13(5) :463-465.
33. Menotti D,Najiman L,Facon J et al. Multi-histogram equalization methods for contrast enhancement and brightness preserving. IEEE Transactions on Consumer Electronics.2007, 53 (3):1186-1194.
34.Grundland M,Dodgson N A. Interactive contrast enhancement by histogram warping[C] .International Conference on Computer Vision and Graphics. 2004,32 :832-838.
35.王保平,刘生虎,范九伦.基于模糊熵的自适应图像多层次模糊增强算法[J].电子学报.2005 ,33(4):730-734.
36.刘博,胡正平,王成儒.基于模糊松弛迭代的分层图像增强算法[J].光学技术.2009,35(1):134-137.
37.Morlet.Wavelet Transform for earthquake analysis[J].Signal Processing.1980, 7(2):1055-1061.
38.Arenodo,G.Grasseau.Wavelet Transform of Multifractals[J].Physical Review letters.1988, 6(1):2281-2284.
39.Jeon,Y.S.Shin.Wavelet Transform for time-frequency Analysis of the Vibration Signature and its Application[J] . IEEE Trans Image Processing. 1996,6(4):1302-1306.
40.Rose. A new algorithm about wavelet hard thresholding for X-ray image defection . IEEE Trans Image Processing .2006, 15(9):13-19.
41.Puillion.Wavelet Transform for X-ray image defection[J].IEEE Trans Image Processing.2006, 23(1):111-117.
42.钢铁.基于神经网络的智能化超声焊接缺陷模式识别与诊断.硕士论文.哈尔滨:哈尔滨工业大学,1995.
43.张广明.超声无损检测中的时频分析理论及应用研究.西安:西安交通大学1999,91-94.
44.刘镇清.超声无损检测与评价中的信号处理及模式识别[J].无损检测2001,23(6):31-34.
45.王恩亮.X射线无损探伤中的缺陷的自动检测和识别.大连:大连理工大学, 2003:73-77.
46.Kenneth R.Castleman著,朱志刚等译.数字图象处理[M].北京:电子工业出版社,2006:265-267.
47.[法]Stephane Mallat著.杨力华,戴道清,黄文良等译.信号处理的小波导引[M].北京:机械工业出版社,2002:106-109.
48.王彬.熵与信息[M].西安:西北工业大学出版社,1994:154-155.
49.周延武,李保国,宗光华.小波模糊理论在擦窗机器人实时图象检测中的应用[J].机器人.2002,24[5]: 26-30.
50.徐碧.基于小波变换的图象增强方法研究.硕士论文.长春:中国科学院光电技术研究所,1999.
51.薛琴,邹道文,吕珂.基于模平方处理的小波图像增强方法[J].江西师范大学学报.2005, 37(6): 276-279.
52.Bovis,Singh.A algorithm about distribution separation measurement for contrast enhancement[J].IEEE Trans Image Processing.1994, 7(4):253-259.
53.Margagos P.Differential and image processing[C].IEEE Trans Image Processing.1996 ,5(6):922-937.
54.杨述斌.数学形态学在数字图像处理中的应用研究.博士论文.武汉:华中科技大学,2002.
55.Canny.A Computational approach to edge detection[J].IEEE Transon PAMI.1985 ,8(6):679-698.
56.Se`bastien Guillion et al.Adaptive nonlinear filters 2D and 3D image enhancement[J].Signal Processing.1998,6(7):237-240.
57.DericheR.Using Canny’s criteria to derive are cursively implemented optimal edge detection[J].The International Journal of Computer Vision.1987, 12(1):167-187.
58.杨述斌,彭复员.噪声污染图像中的广义形态边缘检测器[J].计算机工程与应用.2002,38(7):91-92.
59.杨述斌,彭复员,张增常.多尺度自适应加权形态边缘检测器[J].华中科技大学学报.2002,30(10):42-45.
60.胡晓辉,吕登峰,李敬文.一种基于形态学多结构元的自适应边缘检测算法[J].计算机应用研究.2009,26(2): 752-755.
61.奕新,朱铁一.二次滤波法提取图像边缘信息方法及应用[J].青岛海洋大学学报.1999,29(1):107-111.
62.苑玮琪,王建军,张宏勋.一种基于梯度极值的边缘检测算法[J].信息与控制.1997,26(2):117-120.
63.Stephane Mallat,WenLiang Huang.Singularity detection and processing with wavelets[J].IEEE Transon Information Theory.1992,38(2):617-643.