基于多分辨分析理论的数字图像融合方法
|
摘要
目前,图像融合技术已广泛应用在军事、遥感、机器人、医学处理以及计算机视觉等领域中。传统小波变换能够有效表示信号的点奇异性,但不能较好地表示图像结构中的直线和曲线奇异性；近年来,针对传统小波变换的局限性,为了更好地检测、表示和处理图像等高维空间数据,出现了多尺度几何分析理论。Curvelet(曲波)作为一种新的多尺度分析方法,它比小波更加适合分析二维图像中的曲线或直线状边缘特征,而且具有更高的逼近精度和更好的稀疏表达能力。
本文介绍了图像融合处理的概念和几种常用的变换,分析了第二代Curvelet变换的原理和实现算法。由于遥感图像的数据量非常庞大,本文设计了一种基于小波变换的并行图像融合算法,以克服单机处理在计算能力和存储空间上的限制。在MPI环境下进行了模拟实验,在数据量较大时获得良好的加速比和并行效率。
将第二代Curvelet变换引入图像融合领域,通过把低空间分辨率的多光谱图像重采样达到和全色图像一样的尺寸,对全色图像和多光谱图像分别采用Curvelet变换分解,采用一定的规则融合Curvelet变换后的系数,使得融合后的图像既增加了全色图像的空间细节信息,也保持了多光谱图像的光谱特征。然后对Curvelet算法采用一系列优化措施,提高算法执行效率。实验结果表明,基于Curvelet变换的遥感图像融合算法具有良好的性能和效率,同时具有较快的运行速度。
Image fusion has been widely used in many fields such as military application, remote sensing, robot engineering, medical imaging, computer vision, and so on. The traditional wavelet transform is good at isolating the discontinuities at points, but cannot effectively represent the line discontinuities and the curve discontinuities. To overcome the disadvantages of the wavelets in image analysis, a serial of multiscale geometric analysis(MGA)tools ate proposed in recent years. Curvelet, as a revolutionary method of multiscale geometric analysis, is better suitable in analyzing the property of 2D curves and straight line like edge in comparison to the conventional wavelet method. It is revolutionary also in the sense of higher accuracy close to limits and better sparse representations.
This paper introduces image fusion concept and the common transforms followed by a discussion of our second generation curvelet transform, the principle and the way of calculation. Due to the enormous data processing of remote sensing image, a new parallel algorithm based on wavelet transform is proposed and completed under MPI(Message Processing Interface)circumstance, the considerable acceleration ratio and high efficiency, which is improved compared with only one processor, have been acquired.
A novel fusion algorithm for remote sensing image based on the second generation Curvelet transform is proposed. The low-resolution MS bands are resampled to the fine scale of the panchromatic (PAN) image. The PAN image and MS bands are decomposed by the Curvelet transform. Then, some fusion rules are employed to obtain the Curvelet coefficients of the fusion image, so that the proposed algorithm can not only improve the spatial quality of the fused MS image effectively, but also make the fused images in little spectrum distortion. Meanwhile, various optimization techniques are used on a form of the Curvelet transform in order to improve the execution speed and efficiency of the algorithm. Experiments are carried out on very-high-resolution MS+PAN images acquired by the Ikonos satellite systems, experiments results indicate the proposed Curvelet-based fusion method performs slightly better than wavelet-base fusion method, and have relatively fast execution speed.
本文介绍了图像融合处理的概念和几种常用的变换,分析了第二代Curvelet变换的原理和实现算法。由于遥感图像的数据量非常庞大,本文设计了一种基于小波变换的并行图像融合算法,以克服单机处理在计算能力和存储空间上的限制。在MPI环境下进行了模拟实验,在数据量较大时获得良好的加速比和并行效率。
将第二代Curvelet变换引入图像融合领域,通过把低空间分辨率的多光谱图像重采样达到和全色图像一样的尺寸,对全色图像和多光谱图像分别采用Curvelet变换分解,采用一定的规则融合Curvelet变换后的系数,使得融合后的图像既增加了全色图像的空间细节信息,也保持了多光谱图像的光谱特征。然后对Curvelet算法采用一系列优化措施,提高算法执行效率。实验结果表明,基于Curvelet变换的遥感图像融合算法具有良好的性能和效率,同时具有较快的运行速度。
Image fusion has been widely used in many fields such as military application, remote sensing, robot engineering, medical imaging, computer vision, and so on. The traditional wavelet transform is good at isolating the discontinuities at points, but cannot effectively represent the line discontinuities and the curve discontinuities. To overcome the disadvantages of the wavelets in image analysis, a serial of multiscale geometric analysis(MGA)tools ate proposed in recent years. Curvelet, as a revolutionary method of multiscale geometric analysis, is better suitable in analyzing the property of 2D curves and straight line like edge in comparison to the conventional wavelet method. It is revolutionary also in the sense of higher accuracy close to limits and better sparse representations.
This paper introduces image fusion concept and the common transforms followed by a discussion of our second generation curvelet transform, the principle and the way of calculation. Due to the enormous data processing of remote sensing image, a new parallel algorithm based on wavelet transform is proposed and completed under MPI(Message Processing Interface)circumstance, the considerable acceleration ratio and high efficiency, which is improved compared with only one processor, have been acquired.
A novel fusion algorithm for remote sensing image based on the second generation Curvelet transform is proposed. The low-resolution MS bands are resampled to the fine scale of the panchromatic (PAN) image. The PAN image and MS bands are decomposed by the Curvelet transform. Then, some fusion rules are employed to obtain the Curvelet coefficients of the fusion image, so that the proposed algorithm can not only improve the spatial quality of the fused MS image effectively, but also make the fused images in little spectrum distortion. Meanwhile, various optimization techniques are used on a form of the Curvelet transform in order to improve the execution speed and efficiency of the algorithm. Experiments are carried out on very-high-resolution MS+PAN images acquired by the Ikonos satellite systems, experiments results indicate the proposed Curvelet-based fusion method performs slightly better than wavelet-base fusion method, and have relatively fast execution speed.
引文
[1]G Piella. A general framework for multiresolution image fusion:from pixels to regions, Information Fusion,2003(4):259-280.
[2]焦李成,谭山.图像的多尺度几何分析：回顾和展望[J]电子学报,2003,31(12A)：1975-1981
[3]E J Candes, D L Donoho. Curvelet-A surprisingly effective nonadaptive representation for objects with edges. Curve and Surface Fitting. Vanderbilt Univ Press,1999.34-39
[4]E J Candes, L Demanet, D L Donoho, L Ying. Fast Discrete Curvelet Transforms. Technical Report. CalTech,2005:178-182
[5]M Frigo. FFTW Software Package. Matteo Frigo and Massachusetts Institue of Technology,2006
[6]Mallat. A Wavelet Tour of Signal Process. Academic Press San Diego California,1998.99-105
[7]E J Candes, D L Donoho. New tight frames of Curvelets and optimal represetations of objects with piecewise-C2 singularities,2004:142-145
[8]Brian Eriksson. the very fast Curvelet transform,2003
[9]L Ying. CurveLab 2.0. California Institute of Technology,2005
[10]Intel. IA-32 Intel Architecture Software Developer's Manual. Volume 1:Basic Architecture,2006:98-101
[11]倪林.一种更适合图像处理的多尺度变换-Curvelet变换.计算机工程与应用,2004,4028：21-26
[12]张强,郭宝龙.一种基于Curvelet变换多传感器图像融合算法.光电子,200617(9)：1123-1127
[13]Zhong Zhang, Rick.S.Blum. Image fusion for Digital Camera Application,1998
[14]张强,郭宝龙.应用第二代Curvelet变换的遥感图像融合.光学精密工程,2007：101-104
[15]Gonzalo Pajares, J.M.Cruz. A Wavelet-based image fusion tutorial,2004
[16]陶冰洁,王敬儒,许俊平.基于小波分析的不同融合规则的图像融合.红外技术,2006：210-213
[17]P J Burt, E H Adelson. The Laplacian Pyramid as a Compact Image Code. IEEE Transactions on Communications,1983:72-75
[18]T Ranchin, L Wald. Fusion of high spatial and spectral resolution images:the
ARSIS concept and its implementation. Photogram and Remote Sensing,2000
[19]H Li, B S Manjunath, S K Mitra. Multisensor image fusion using the wavelet transform. Graphical Models Image Process,1995:134-136
[20]Myungjin choi, Rae Young Kim, Myeong-Ryong Nam. Fusion of multispectral and panchromatic Satellite images using the Curvelet transform. Geoscience and Remote Sensing Letters IEEE,2005:241-245
[21]Garelli A, Nencini F, Alparone. Multiresolution fusion of multispectral and panchromatic images through the curvelet transform. IEEE,2005:345-348
[22]M N Do. Directional Multiresolution Image represetations. PhD thesis,2001
[23]M N Do, M Vetterli. Contourlets in Beyond Wavelets. Academic Press,2003
[24]J L Starck, E J Candes, D L Donoho. The Curvelet transform for image denoising. IEEE Trans,2002:156-158
[25]Zhong Zhang. Rich.S.Blum. A Hybrid Image Registration Technique for A Digital Camera Image Fusion Application. Information Fusion,2001:34-36
[26]Donoho D L. Orthonormal ridgelets and linear singularities. SIAM J. Math Anal, 2000,31(5):1062-1099
[27]Hall D L, Llinas J. An introduction to multisensory data fusion. In Proceeding of IEEE,1997,85(1):6-23
[28]Li H, Manjunath B S, Mitra S K. Multisensor image fusion using the wavelet transform. Computer Vision, Graphics, and Image Processing:Graphical Models and Image Processing,1993:235-245
[29]Houzelle S, Giraudon G. Contribution to multisensor fusion formaliza tion. Robotics and Autonomous Systems,1994,13:69-85
[30]Hall D, Mathematical technique in multisensory data fusion. Artech House, Boston, London,1992:148-151
[31]Klein L A. Sensor and data fusion concepts and applications. SPIE Optical Engineering Press, USA,1994:20-24
[32]Thomopoulos S C A. Sensor integration and data fusion. Journal of Robotic Systems,1990,7:337-372
[33]AkermanⅢ A. Pyramid techniques for multisensory fusion. In Proc. SPIE,1992, 1828:124-131
[34]Zhou Y.T. Multisensor image fusion. In Proc. IEEE Int. Conf. Image Processing, IEEE'94, Austin, TX:193-197
[35]Valdimir.S.Petrovic, Costas.S.Xydeas. Gradient-Based Multiresolution Image Fusion. IEEE Image Processing,2004,13(2):228-237
[36]S Li, J T Kwok, Y Wang. Using the discrete wavelet frame transform to merge Landsat TM and SPOT panchromatic images, Information Fusion 2002, 3(1):259-280
[37]D.A.Yocky. Artifacts in wavelet image merging. Optical Engineering,1996, 35(7):2094-2101
[38]M Gonzales Audicana, J L Saleta, R Garcia Catalan, R Garcia. Fusion of multispectral and panchromatic images using improved IHS and PCA mergers based on wavelet decomposition. IEEE Transaction On Geosciences and Remote Sensing,2004,42(6):1291-1299
[39]J L Starck, F Murtagh, E J Candes, D L Donoho. Gray and color image contrast enhancement by the curvelet transform. IEEE Transactions on Image Processing, 2003,12(6):706-717
[40]A Garzelli, F Nencini. Interband structure modeling for Pan-sharpening of very high resolution multispectral images. Information Fusion,2005,6(3):213-224
[41]J L Starck, F Murtagh. Image restoration with noise suppression using the wavelet transform. Astronomy and Astrophysics,1994,288:342-350
[42]Y Chibani, A Houacine. The joint use of IHS transform and redundant wavelet decomposition for fusing multispectral and panchromatic images. International Journal of Remote Sensing,2002,23(18):3821-3833
[43]P S Chavez Jr, S C Slides, J A Anderson. Comparison of three different methods to merge multiresolution and multispectral data:Landsat TM and SPOT panchromatic. Photogrammetric Engineering and Remote Sensing,1991, 57(3):295-303
[44]G Piella, H Heijmans. A new quality metric for image fusion. Proceedings of the IEEE International Conference on Image Processing, vol.III/IV, 2003:173-176
[45]Y Zhang. Understanding image fusion. Photogrammetric Engineering and Remote Sensing,2004,70(6):657-661
[46]T M Tu, S C Su, H C Shyu. A new look at IHS like image fusion methods. Information Fusion,2001,2(3):177-186
[47]Chibani Y, Houacine A. Redundant versus orthogonal wavelet decomposition for multisensory image fusion. Pattern Recognition,2003,36:879-887
[48]Simone G, Farina A. Image fusion techniques for remote sensing applications Information Fusion,2002,3:3-15
[49]Zhu Shulong, Zhang ZhangKui. Remote Sensing Image Acquisition and Analysis. Beijing:Science Press,2000:48-50
[50]M T Eismann, R C Hardie. Hyperspectral resolution enhancement using high-resolution multispectral imagery with arbitrary response functions. IEEE Transactions On Geoscience and Remote Sensing,2005,43(3):455-465
[2]焦李成,谭山.图像的多尺度几何分析：回顾和展望[J]电子学报,2003,31(12A)：1975-1981
[3]E J Candes, D L Donoho. Curvelet-A surprisingly effective nonadaptive representation for objects with edges. Curve and Surface Fitting. Vanderbilt Univ Press,1999.34-39
[4]E J Candes, L Demanet, D L Donoho, L Ying. Fast Discrete Curvelet Transforms. Technical Report. CalTech,2005:178-182
[5]M Frigo. FFTW Software Package. Matteo Frigo and Massachusetts Institue of Technology,2006
[6]Mallat. A Wavelet Tour of Signal Process. Academic Press San Diego California,1998.99-105
[7]E J Candes, D L Donoho. New tight frames of Curvelets and optimal represetations of objects with piecewise-C2 singularities,2004:142-145
[8]Brian Eriksson. the very fast Curvelet transform,2003
[9]L Ying. CurveLab 2.0. California Institute of Technology,2005
[10]Intel. IA-32 Intel Architecture Software Developer's Manual. Volume 1:Basic Architecture,2006:98-101
[11]倪林.一种更适合图像处理的多尺度变换-Curvelet变换.计算机工程与应用,2004,4028：21-26
[12]张强,郭宝龙.一种基于Curvelet变换多传感器图像融合算法.光电子,200617(9)：1123-1127
[13]Zhong Zhang, Rick.S.Blum. Image fusion for Digital Camera Application,1998
[14]张强,郭宝龙.应用第二代Curvelet变换的遥感图像融合.光学精密工程,2007：101-104
[15]Gonzalo Pajares, J.M.Cruz. A Wavelet-based image fusion tutorial,2004
[16]陶冰洁,王敬儒,许俊平.基于小波分析的不同融合规则的图像融合.红外技术,2006：210-213
[17]P J Burt, E H Adelson. The Laplacian Pyramid as a Compact Image Code. IEEE Transactions on Communications,1983:72-75
[18]T Ranchin, L Wald. Fusion of high spatial and spectral resolution images:the
ARSIS concept and its implementation. Photogram and Remote Sensing,2000
[19]H Li, B S Manjunath, S K Mitra. Multisensor image fusion using the wavelet transform. Graphical Models Image Process,1995:134-136
[20]Myungjin choi, Rae Young Kim, Myeong-Ryong Nam. Fusion of multispectral and panchromatic Satellite images using the Curvelet transform. Geoscience and Remote Sensing Letters IEEE,2005:241-245
[21]Garelli A, Nencini F, Alparone. Multiresolution fusion of multispectral and panchromatic images through the curvelet transform. IEEE,2005:345-348
[22]M N Do. Directional Multiresolution Image represetations. PhD thesis,2001
[23]M N Do, M Vetterli. Contourlets in Beyond Wavelets. Academic Press,2003
[24]J L Starck, E J Candes, D L Donoho. The Curvelet transform for image denoising. IEEE Trans,2002:156-158
[25]Zhong Zhang. Rich.S.Blum. A Hybrid Image Registration Technique for A Digital Camera Image Fusion Application. Information Fusion,2001:34-36
[26]Donoho D L. Orthonormal ridgelets and linear singularities. SIAM J. Math Anal, 2000,31(5):1062-1099
[27]Hall D L, Llinas J. An introduction to multisensory data fusion. In Proceeding of IEEE,1997,85(1):6-23
[28]Li H, Manjunath B S, Mitra S K. Multisensor image fusion using the wavelet transform. Computer Vision, Graphics, and Image Processing:Graphical Models and Image Processing,1993:235-245
[29]Houzelle S, Giraudon G. Contribution to multisensor fusion formaliza tion. Robotics and Autonomous Systems,1994,13:69-85
[30]Hall D, Mathematical technique in multisensory data fusion. Artech House, Boston, London,1992:148-151
[31]Klein L A. Sensor and data fusion concepts and applications. SPIE Optical Engineering Press, USA,1994:20-24
[32]Thomopoulos S C A. Sensor integration and data fusion. Journal of Robotic Systems,1990,7:337-372
[33]AkermanⅢ A. Pyramid techniques for multisensory fusion. In Proc. SPIE,1992, 1828:124-131
[34]Zhou Y.T. Multisensor image fusion. In Proc. IEEE Int. Conf. Image Processing, IEEE'94, Austin, TX:193-197
[35]Valdimir.S.Petrovic, Costas.S.Xydeas. Gradient-Based Multiresolution Image Fusion. IEEE Image Processing,2004,13(2):228-237
[36]S Li, J T Kwok, Y Wang. Using the discrete wavelet frame transform to merge Landsat TM and SPOT panchromatic images, Information Fusion 2002, 3(1):259-280
[37]D.A.Yocky. Artifacts in wavelet image merging. Optical Engineering,1996, 35(7):2094-2101
[38]M Gonzales Audicana, J L Saleta, R Garcia Catalan, R Garcia. Fusion of multispectral and panchromatic images using improved IHS and PCA mergers based on wavelet decomposition. IEEE Transaction On Geosciences and Remote Sensing,2004,42(6):1291-1299
[39]J L Starck, F Murtagh, E J Candes, D L Donoho. Gray and color image contrast enhancement by the curvelet transform. IEEE Transactions on Image Processing, 2003,12(6):706-717
[40]A Garzelli, F Nencini. Interband structure modeling for Pan-sharpening of very high resolution multispectral images. Information Fusion,2005,6(3):213-224
[41]J L Starck, F Murtagh. Image restoration with noise suppression using the wavelet transform. Astronomy and Astrophysics,1994,288:342-350
[42]Y Chibani, A Houacine. The joint use of IHS transform and redundant wavelet decomposition for fusing multispectral and panchromatic images. International Journal of Remote Sensing,2002,23(18):3821-3833
[43]P S Chavez Jr, S C Slides, J A Anderson. Comparison of three different methods to merge multiresolution and multispectral data:Landsat TM and SPOT panchromatic. Photogrammetric Engineering and Remote Sensing,1991, 57(3):295-303
[44]G Piella, H Heijmans. A new quality metric for image fusion. Proceedings of the IEEE International Conference on Image Processing, vol.III/IV, 2003:173-176
[45]Y Zhang. Understanding image fusion. Photogrammetric Engineering and Remote Sensing,2004,70(6):657-661
[46]T M Tu, S C Su, H C Shyu. A new look at IHS like image fusion methods. Information Fusion,2001,2(3):177-186
[47]Chibani Y, Houacine A. Redundant versus orthogonal wavelet decomposition for multisensory image fusion. Pattern Recognition,2003,36:879-887
[48]Simone G, Farina A. Image fusion techniques for remote sensing applications Information Fusion,2002,3:3-15
[49]Zhu Shulong, Zhang ZhangKui. Remote Sensing Image Acquisition and Analysis. Beijing:Science Press,2000:48-50
[50]M T Eismann, R C Hardie. Hyperspectral resolution enhancement using high-resolution multispectral imagery with arbitrary response functions. IEEE Transactions On Geoscience and Remote Sensing,2005,43(3):455-465