一种多光谱红外舰船图像多级融合方法
|
摘要
针对复杂背景下多光谱红外舰船目标图像融合问题,根据不同谱段中目标与背景,杂波光谱特性的差异性,结合杂波产生随机性特点和不同谱段信息的相关性,提出了一种红外多谱段图像多级融合算法。首先根据不同谱段信息相互间关系进行融合分级,然后对同一级中图像分别进行NSCT分解,再采用基于局部能量比和基于局部能量加权的方法进行低频系数融合,采用基于尺度方差选大法进行高频系数融合,最后通过NSCT重构得到融合图像。实验结果表明,本文提出的多级融合算法有效保留目标细节信息,增加信息量,利于目标检测。
In order to solve the multi-spectrum image fusion problem of infrared ship targets under complex background,according to the difference of the spectral characteristics of the target and background in different spectral segments and clutter,combined with the randomness of clutter generation and the correlation of different spectral segment information,an infrared multi-spectral image multi-level fusion algorithm is proposed. Firstly,according to the relationship between the information of different spectral segments,the fusion and classification are performed. Then,the images in the same level are separately decomposed by NSCT,and then the low-frequency coefficient fusion is performed by the methods based on local energy ratio and local energy weighting,the high-frequency coefficient fusion is performed using a method based on scale variance selection,and fusion images are finally obtained through NSCT reconstruction. Experimental results show that the multi-spectral infrared ship image multi-level fusion algorithm proposed in this paper effectively preserves the target details,increases the amount of information,and facilitates target detection.
In order to solve the multi-spectrum image fusion problem of infrared ship targets under complex background,according to the difference of the spectral characteristics of the target and background in different spectral segments and clutter,combined with the randomness of clutter generation and the correlation of different spectral segment information,an infrared multi-spectral image multi-level fusion algorithm is proposed. Firstly,according to the relationship between the information of different spectral segments,the fusion and classification are performed. Then,the images in the same level are separately decomposed by NSCT,and then the low-frequency coefficient fusion is performed by the methods based on local energy ratio and local energy weighting,the high-frequency coefficient fusion is performed using a method based on scale variance selection,and fusion images are finally obtained through NSCT reconstruction. Experimental results show that the multi-spectral infrared ship image multi-level fusion algorithm proposed in this paper effectively preserves the target details,increases the amount of information,and facilitates target detection.
引文
[1] Hoff L E,Evans J R,Bunney L E.Detection of targets in terrain clutter by using multispectral infrared image processing[J].Proc Spie,1990,1481:54.
[2] Yu X,Reed I S,Stocker A D.Comparative performance analysis of adaptive multispectral detectors[M].IEEE Press,1993:2639-2656.
[3] Margalit A,Reed I S,Gagliardi R M.Adaptive optical target detection using correlated images[J].IEEE Transactions on Aerospace & Electronic Systems,1985,AES-21(3):394-405.
[4] Miller M I,Fuhrmann D R.Maximum-likelihood narrow-band direction finding and the EM algorithm[J].IEEE Transactions on Acoustics Speech & Signal Processing,1990,38(9):1560-1577.
[5] WU Chunfeng,ZHANG Wei,CONG Mingxi,et al.Automatic target recognition algorithm based on infrared multispectral image correlation[J].Journal of Infrared and Millimeter Waves,2003,22(4):265-268.(in Chinese)武春风,张伟,丛明煜,等.基于红外多光谱图像相关性的自动目标识别算法[J].红外与毫米波学报,2003,22(4):265-268.
[6] GUO Min,JIANG Aimin,CAO Mei.A new method for nonlinear enhancement of infrared image based on nonsubsampled contourlet transform[J].Journal of Computers and Modernization,2017,(3):76-79.(in Chinese)郭珉,蒋爱民,曹美.基于非下采样Contourlet变换的红外图像非线性增强新方法[J].计算机与现代化,2017,(3):76-79.
[7] Da C A,Zhou J,Do M N.The nonsubsampled contourlet transform:theory,design,and applications[J].IEEE Transactions on Image Processing,2006,15(10):3089-3101.
[8] GUO Ming,FU Zheng,YAN Xiaoliang.Infrared and visible light image fusion algorithm based on local energy in NSCT domain[J].Infrared and Laser Engineering,2012,41(8):2229-2235.(in Chinese)郭明,符拯,奚晓梁.基于局部能量的NSCT域红外与可见光图像融合算法[J].红外与激光工程,2012,41(8):2229-2235.
[9] YE Chuanqi,WANG Baoshu,MIAO Qiguang.A multispectral and high resolution image fusion algorithm based on region in NSCT domain[J].Acta Optica Sinica,2009,29(5):1240-1247.(in Chinese)叶传奇,王宝树,苗启广.一种基于区域的NSCT域多光谱与高分辨率图像融合算法[J].光学学报,2009,29(5):1240-1247.
[10] WANG Leshan,LI Huafeng,A Zhu-La-qie.Contrast image fusion in non-sampling Contourlet domain[J].Computer Engineering and Applications,2010,46(35):185-187.(in Chinese)王乐山,李华锋,阿主拉且.非采样Contourlet域内的区域对比度图像融合[J].计算机工程与应用,2010,46(35):185-187.
[11] ZHANG Jinglei,ZHAO Eying.Infrared and visible light image fusion method based on NSCT[J].LASER & Infrared,2013,43(3):319-323.(in Chinese)张惊雷,赵俄英.基于NSCT的红外与可见光图像融合方法[J].激光与红外,2013,43(3):319-323.
[12] ZHU Yahui.Research on similarity of infrared and visible fusion quality indexes based on cluster analysis[J].Computer Systems Applications,2018,27(2):216-222.(in Chinese)朱亚辉.聚类分析的红外与可见光融合质量指标相似性研究[J].计算机系统应用,2018,27(2):216-222.
[13] ZHANG Xiaoli,LI Xiongfei,LI Jun.Correlation analysis and performance evaluation of fusion image quality evaluation index[J].Acta Automatica Sinica,2014,40(2):306-315.(in Chinese)张小利,李雄飞,李军.融合图像质量评价指标的相关性分析及性能评估[J].自动化学报,2014,40(2):306-315.
[14] Liu Z,Blasch E,Xue Z,et al.Objective Assessment of multiresolution image fusion algorithms for context enhancement in night vision:a comparative study[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2011,34(1):94-109.
[2] Yu X,Reed I S,Stocker A D.Comparative performance analysis of adaptive multispectral detectors[M].IEEE Press,1993:2639-2656.
[3] Margalit A,Reed I S,Gagliardi R M.Adaptive optical target detection using correlated images[J].IEEE Transactions on Aerospace & Electronic Systems,1985,AES-21(3):394-405.
[4] Miller M I,Fuhrmann D R.Maximum-likelihood narrow-band direction finding and the EM algorithm[J].IEEE Transactions on Acoustics Speech & Signal Processing,1990,38(9):1560-1577.
[5] WU Chunfeng,ZHANG Wei,CONG Mingxi,et al.Automatic target recognition algorithm based on infrared multispectral image correlation[J].Journal of Infrared and Millimeter Waves,2003,22(4):265-268.(in Chinese)武春风,张伟,丛明煜,等.基于红外多光谱图像相关性的自动目标识别算法[J].红外与毫米波学报,2003,22(4):265-268.
[6] GUO Min,JIANG Aimin,CAO Mei.A new method for nonlinear enhancement of infrared image based on nonsubsampled contourlet transform[J].Journal of Computers and Modernization,2017,(3):76-79.(in Chinese)郭珉,蒋爱民,曹美.基于非下采样Contourlet变换的红外图像非线性增强新方法[J].计算机与现代化,2017,(3):76-79.
[7] Da C A,Zhou J,Do M N.The nonsubsampled contourlet transform:theory,design,and applications[J].IEEE Transactions on Image Processing,2006,15(10):3089-3101.
[8] GUO Ming,FU Zheng,YAN Xiaoliang.Infrared and visible light image fusion algorithm based on local energy in NSCT domain[J].Infrared and Laser Engineering,2012,41(8):2229-2235.(in Chinese)郭明,符拯,奚晓梁.基于局部能量的NSCT域红外与可见光图像融合算法[J].红外与激光工程,2012,41(8):2229-2235.
[9] YE Chuanqi,WANG Baoshu,MIAO Qiguang.A multispectral and high resolution image fusion algorithm based on region in NSCT domain[J].Acta Optica Sinica,2009,29(5):1240-1247.(in Chinese)叶传奇,王宝树,苗启广.一种基于区域的NSCT域多光谱与高分辨率图像融合算法[J].光学学报,2009,29(5):1240-1247.
[10] WANG Leshan,LI Huafeng,A Zhu-La-qie.Contrast image fusion in non-sampling Contourlet domain[J].Computer Engineering and Applications,2010,46(35):185-187.(in Chinese)王乐山,李华锋,阿主拉且.非采样Contourlet域内的区域对比度图像融合[J].计算机工程与应用,2010,46(35):185-187.
[11] ZHANG Jinglei,ZHAO Eying.Infrared and visible light image fusion method based on NSCT[J].LASER & Infrared,2013,43(3):319-323.(in Chinese)张惊雷,赵俄英.基于NSCT的红外与可见光图像融合方法[J].激光与红外,2013,43(3):319-323.
[12] ZHU Yahui.Research on similarity of infrared and visible fusion quality indexes based on cluster analysis[J].Computer Systems Applications,2018,27(2):216-222.(in Chinese)朱亚辉.聚类分析的红外与可见光融合质量指标相似性研究[J].计算机系统应用,2018,27(2):216-222.
[13] ZHANG Xiaoli,LI Xiongfei,LI Jun.Correlation analysis and performance evaluation of fusion image quality evaluation index[J].Acta Automatica Sinica,2014,40(2):306-315.(in Chinese)张小利,李雄飞,李军.融合图像质量评价指标的相关性分析及性能评估[J].自动化学报,2014,40(2):306-315.
[14] Liu Z,Blasch E,Xue Z,et al.Objective Assessment of multiresolution image fusion algorithms for context enhancement in night vision:a comparative study[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2011,34(1):94-109.