基于小波变换的图像融合技术研究
|
摘要
图像融合技术是由信息融合技术发展而来的,其根本方法就是对多个传感器采集的原始图像,使用特定的融和模型,融合生成一幅新的融和图像。融合后的图像不仅具有原始图像的大部分信息,而且更符合人眼的视觉习惯和机器感知,能对特定场景进行更全面、更精确地描述。近年以来,图像融合技术己经成为了一个重要研究领域,应用十分广泛。
图像融合按层次分类,可分为像素级融合、特征级融合、决策级融合,其中像素级融合是基础。本文主要探讨像素级图像融合技术。
本文根据图像融合技术的研究背景与现状,简要介绍了图像融合中的一些基础问题,并对像素级图像融合方法展开了研究,主要完成的工作如下：
(1)本文简要说明图像融合技术的应用范围、研究现状、发展前景和困难,指出图像融合的意义和重要性,并在此基础上,对图像融合的预处理阶段的图像增强和图像配准操作做了介绍。
(2)阐述了图像融合的步骤,探讨了图像融合的主观、客观质量评价标准,并对评价指标的选取问题,给出了简要说明；综合分析、介绍了小波变换理论及其在图像融合方面的应用。
(3)对比研究了常用的图像融合方法,根据常用融合方法做了大量仿真试验；对照仿真试验结果,指出了其中的不足,在此基础上对基于区域能量融合算法的小波融合算法进行了改进。该方法利用小波分解,将源图像分解成低频和高频两个部分,对低频部分,采用了自适应加权系数；高频部分采用局部能量与加权相结合的融合方法。利用左、右聚焦图像和红外图像、可见光图像进行了仿真实验,并运用图像客观评价指标验证该方法的可行性
Image fusion is evolved by information fusion, such process generates a single image which contains a more accurate description of the scene than any of the individual source images. This fused image should be more useful for human visual or machine perception, and it could be more accurate and more comprehensive description of the certain image scene through the processing of redundant information and complementary information of the source image. In recent years, image fusion has become a very important research field of many subjects and has been widely used.
According to hierarchical classification, image fusion falls into the following three categories:pixel level fusion, feature level fusion, and decision level fusion. The basis is pixel level, and pixel level fusion is the object of study in this paper.
This paper according to the image fusion technology's background and status, gave some of the basic issues on the image fusion, and discuss the methods of pixel-level image fusion, the main work is as follows:
(1) This paper gives a brief description of the image fusion technology's range of applications, study of the current situation and development prospects and difficulties, and pointed out that images the meaning and importance, and on this basis, the images of the pre-processing stage of image enhancement and image registration method.
(2) Explain the steps and comparative study of image fusion methods commonly used discussion on image fusion of subjective and objective quality evaluation criteria and the selection of evaluation index, and gives a brief description; Comprehensive analysis, introduction to the theory and its application in image fusion based on wavelet transform.
(3) According to the common fusion methods to do a lot of simulation test, control simulation test results, pointed out that the lack of discussion on image fusion, and improved a method which based on regional local average gradient fusion with Wavelet Decomposition method. The method using Wavelet Decomposition of the original image for low frequency and high frequency sections, for low frequency section, using a weighted coefficient; high frequency part of the use of "local average gradient" fusion methods. Using visible light, infrared images left and focus image focus image and right carried out simulation experiments, using objective evaluation to verify the feasibility of the method.
图像融合按层次分类,可分为像素级融合、特征级融合、决策级融合,其中像素级融合是基础。本文主要探讨像素级图像融合技术。
本文根据图像融合技术的研究背景与现状,简要介绍了图像融合中的一些基础问题,并对像素级图像融合方法展开了研究,主要完成的工作如下：
(1)本文简要说明图像融合技术的应用范围、研究现状、发展前景和困难,指出图像融合的意义和重要性,并在此基础上,对图像融合的预处理阶段的图像增强和图像配准操作做了介绍。
(2)阐述了图像融合的步骤,探讨了图像融合的主观、客观质量评价标准,并对评价指标的选取问题,给出了简要说明；综合分析、介绍了小波变换理论及其在图像融合方面的应用。
(3)对比研究了常用的图像融合方法,根据常用融合方法做了大量仿真试验；对照仿真试验结果,指出了其中的不足,在此基础上对基于区域能量融合算法的小波融合算法进行了改进。该方法利用小波分解,将源图像分解成低频和高频两个部分,对低频部分,采用了自适应加权系数；高频部分采用局部能量与加权相结合的融合方法。利用左、右聚焦图像和红外图像、可见光图像进行了仿真实验,并运用图像客观评价指标验证该方法的可行性
Image fusion is evolved by information fusion, such process generates a single image which contains a more accurate description of the scene than any of the individual source images. This fused image should be more useful for human visual or machine perception, and it could be more accurate and more comprehensive description of the certain image scene through the processing of redundant information and complementary information of the source image. In recent years, image fusion has become a very important research field of many subjects and has been widely used.
According to hierarchical classification, image fusion falls into the following three categories:pixel level fusion, feature level fusion, and decision level fusion. The basis is pixel level, and pixel level fusion is the object of study in this paper.
This paper according to the image fusion technology's background and status, gave some of the basic issues on the image fusion, and discuss the methods of pixel-level image fusion, the main work is as follows:
(1) This paper gives a brief description of the image fusion technology's range of applications, study of the current situation and development prospects and difficulties, and pointed out that images the meaning and importance, and on this basis, the images of the pre-processing stage of image enhancement and image registration method.
(2) Explain the steps and comparative study of image fusion methods commonly used discussion on image fusion of subjective and objective quality evaluation criteria and the selection of evaluation index, and gives a brief description; Comprehensive analysis, introduction to the theory and its application in image fusion based on wavelet transform.
(3) According to the common fusion methods to do a lot of simulation test, control simulation test results, pointed out that the lack of discussion on image fusion, and improved a method which based on regional local average gradient fusion with Wavelet Decomposition method. The method using Wavelet Decomposition of the original image for low frequency and high frequency sections, for low frequency section, using a weighted coefficient; high frequency part of the use of "local average gradient" fusion methods. Using visible light, infrared images left and focus image focus image and right carried out simulation experiments, using objective evaluation to verify the feasibility of the method.
引文
[1]宋瑞晶.基于多尺度图像融合方法研究[D].重庆大学硕士学位论文.2010.
[2]D.L.Hall,J.Llinas.An introduction to multi-sensor fusion[C].Proceedings of the IEEE.1997, 85(1):6-23.
[3]覃征,鲍复民.数字图像融合[M].西安大学出版社.2004.07.
[4]闫敬文.数字图像处理[M].国防科技出版社.2008.02.
[5]赵书兰MATLAB R2008数字图像处理与分析实例教程[M].化学工业出版社.2009.
[6]黄红林.基于平均梯度和小波多分辨率分析的图像融合算法研究[D].武汉科技大学硕士学位论文.2006.
[7]张德丰等.数字图像处理[M].机械工业出版社.2009.1.
[8]张强,王正林.精通MATLAB图像处理[M].电子工业出版社.2009.6.
[9]王卫卫等.小波域多聚焦图像融合算法[J].系统工程与电子技术.2004,26(5)：668-670.
[10]何东健.数字图像处理[M].西安电子科技大学出版社.2004.
[11]罗忠.多源遥感数据融合的现状[J].测试技术学报.1999,13(1)：32-38.
[12]夏明革,何友,唐小明等.多传感器图像融合综述[J].电光与控制.2002,9(4)：1-7.
[13]李树涛,王耀南,龚理专.多聚焦图像融合中最佳小波分解层数的选取[J].系统工程与电子技术.2002,24(6)：45-48.
[14]姜庆娟,谭景信.像素级图像融合方法与选择[J].计算机工程与应用.2003,39(25)：27-120.
[15]成礼智,王红霞,罗勇.小波的理论与应用[M].科学出版社.2006.
[16]Tu TeMing,Su ShunChi,Shyu HsuenChyun,etc.A New Look at IHS-like ImageFusion Methods.Information Fusion.2001,2:177-186.
[17]李希宁.基于多尺度几何分析的图像融合算法研究[D].中国海洋大学硕士学位论文.2010.
[18]Li Cao Wen,Chen Bi,Zhang Yong.A Remote Sensing Image Fusion Method Based on PCA Transform and Wavelet Packet Transform.IEEE Int.Conf.Neural Networks&Signal Processing.Nanjing.China.2003,12:1417-1423.
[19]强赞霞.遥感图像的融合及应用[D].华中科技大学博士学位论文.2005.
[20]刘贵喜.多传感器图像融合方法研究[D].西安电子科技大学硕士学位论文.2001.
[21]求是科技.MATLAB7.0从入门到精通[M].人民邮电出版社.2006.3.
[22]马艳君.基于小波变换的医学图像融合技术研究[D].青岛大学硕士学位论文.2010.
[23]边肇祺,张学工.模式识别(第二版)[M].清华大学出版社.2006.
[24]Laine.A.F,Xuli Zong.A Multiscale Sub-octave Wavelet Transform for Denosing and Enhancement [J].Proceeding of SPIE.1996,2825:238-249.
[25]刘勇.基于遗传算法的红外图像分割研究[D].中南大学硕士学位论文.2009.
[26]王明.乳腺X线计算机辅助诊断关键技术研究[D].上海交通大学硕士学位论文.2009.
[27]毕迎春,王相海.小波基和图像分解层数对不同类型图像EZW算法的性能的影响[J].计算机科学.2006,33(6)：232-246.
[28]龚昌来.一种改进的基于局部能量图像融合算法[J].光电工程.2008,11：106-119.
[29]王卫卫,水鹏朗,宋国乡.小波域多聚焦图像融合算法[J].系统工程与电子技术2004,26(5)：668-670.
[30]朱锡芳.光学图像去云雾方法研究[D].南京理工大学大学博士学位论文.2008.6.
[31]孙颖力.图像融合算法的研究及可重构FPGA[D].西北工业大学硕士学位论文.2007.
[32]钱民,徐向民.乳腺X计算机辅助检测系统研究发展[J].放射科实践.2009,24(1)：100-103.
[33]Lawernee A.Klein.Sensor and Data Fusion Concepts and Applications[J].SPIE Optical Engineering Press,1999,14.
[34]Waxman A.M.,Fay D.A.,Gove A.N.,etal.Color night vision:fusion of intensified visible and thermal IR imagery. Synthetic Vision for Vehicle Guidance and Control [J]. Proceed ings of SPIE.1995,2463:58-68.
[35]徐华楠,刘哲,胡钢Contourlet变换及其在图像去噪中的应用研究[J].计算机应用研究.2009,26(2)：401-405.
[36]梁栋,殷兵,于梅.基于非抽样Contourlet变换的自适应阈值图像增强算法[J].电子学报,2008,26(3)：527-530.
[37]赵晓雷.像素级图像融合技术研究[D].西安科技大学硕士学位论文.2010.
[38]田养军.基于提升小波分解曲波变换的多源遥感图像融合方法研究[D].长安大学博士学位论文.2009.
[2]D.L.Hall,J.Llinas.An introduction to multi-sensor fusion[C].Proceedings of the IEEE.1997, 85(1):6-23.
[3]覃征,鲍复民.数字图像融合[M].西安大学出版社.2004.07.
[4]闫敬文.数字图像处理[M].国防科技出版社.2008.02.
[5]赵书兰MATLAB R2008数字图像处理与分析实例教程[M].化学工业出版社.2009.
[6]黄红林.基于平均梯度和小波多分辨率分析的图像融合算法研究[D].武汉科技大学硕士学位论文.2006.
[7]张德丰等.数字图像处理[M].机械工业出版社.2009.1.
[8]张强,王正林.精通MATLAB图像处理[M].电子工业出版社.2009.6.
[9]王卫卫等.小波域多聚焦图像融合算法[J].系统工程与电子技术.2004,26(5)：668-670.
[10]何东健.数字图像处理[M].西安电子科技大学出版社.2004.
[11]罗忠.多源遥感数据融合的现状[J].测试技术学报.1999,13(1)：32-38.
[12]夏明革,何友,唐小明等.多传感器图像融合综述[J].电光与控制.2002,9(4)：1-7.
[13]李树涛,王耀南,龚理专.多聚焦图像融合中最佳小波分解层数的选取[J].系统工程与电子技术.2002,24(6)：45-48.
[14]姜庆娟,谭景信.像素级图像融合方法与选择[J].计算机工程与应用.2003,39(25)：27-120.
[15]成礼智,王红霞,罗勇.小波的理论与应用[M].科学出版社.2006.
[16]Tu TeMing,Su ShunChi,Shyu HsuenChyun,etc.A New Look at IHS-like ImageFusion Methods.Information Fusion.2001,2:177-186.
[17]李希宁.基于多尺度几何分析的图像融合算法研究[D].中国海洋大学硕士学位论文.2010.
[18]Li Cao Wen,Chen Bi,Zhang Yong.A Remote Sensing Image Fusion Method Based on PCA Transform and Wavelet Packet Transform.IEEE Int.Conf.Neural Networks&Signal Processing.Nanjing.China.2003,12:1417-1423.
[19]强赞霞.遥感图像的融合及应用[D].华中科技大学博士学位论文.2005.
[20]刘贵喜.多传感器图像融合方法研究[D].西安电子科技大学硕士学位论文.2001.
[21]求是科技.MATLAB7.0从入门到精通[M].人民邮电出版社.2006.3.
[22]马艳君.基于小波变换的医学图像融合技术研究[D].青岛大学硕士学位论文.2010.
[23]边肇祺,张学工.模式识别(第二版)[M].清华大学出版社.2006.
[24]Laine.A.F,Xuli Zong.A Multiscale Sub-octave Wavelet Transform for Denosing and Enhancement [J].Proceeding of SPIE.1996,2825:238-249.
[25]刘勇.基于遗传算法的红外图像分割研究[D].中南大学硕士学位论文.2009.
[26]王明.乳腺X线计算机辅助诊断关键技术研究[D].上海交通大学硕士学位论文.2009.
[27]毕迎春,王相海.小波基和图像分解层数对不同类型图像EZW算法的性能的影响[J].计算机科学.2006,33(6)：232-246.
[28]龚昌来.一种改进的基于局部能量图像融合算法[J].光电工程.2008,11：106-119.
[29]王卫卫,水鹏朗,宋国乡.小波域多聚焦图像融合算法[J].系统工程与电子技术2004,26(5)：668-670.
[30]朱锡芳.光学图像去云雾方法研究[D].南京理工大学大学博士学位论文.2008.6.
[31]孙颖力.图像融合算法的研究及可重构FPGA[D].西北工业大学硕士学位论文.2007.
[32]钱民,徐向民.乳腺X计算机辅助检测系统研究发展[J].放射科实践.2009,24(1)：100-103.
[33]Lawernee A.Klein.Sensor and Data Fusion Concepts and Applications[J].SPIE Optical Engineering Press,1999,14.
[34]Waxman A.M.,Fay D.A.,Gove A.N.,etal.Color night vision:fusion of intensified visible and thermal IR imagery. Synthetic Vision for Vehicle Guidance and Control [J]. Proceed ings of SPIE.1995,2463:58-68.
[35]徐华楠,刘哲,胡钢Contourlet变换及其在图像去噪中的应用研究[J].计算机应用研究.2009,26(2)：401-405.
[36]梁栋,殷兵,于梅.基于非抽样Contourlet变换的自适应阈值图像增强算法[J].电子学报,2008,26(3)：527-530.
[37]赵晓雷.像素级图像融合技术研究[D].西安科技大学硕士学位论文.2010.
[38]田养军.基于提升小波分解曲波变换的多源遥感图像融合方法研究[D].长安大学博士学位论文.2009.