卷积稀疏表示和邻域特征结合的多聚焦图像融合
|
摘要
稀疏表示的分块处理破环了图像的连续性,导致多聚焦融合图像的清晰测度信息严重丢失。针对上述问题,提出了卷积稀疏表示和邻域特征结合的多聚焦图像融合算法。该算法将非下采样轮廓波变换(NSCT)域低频子图通过高斯滤波分解成基础层和细节层,然后选用交替方向乘子算法(ADMM)求解稀疏系数,完成细节层特征响应系数的融合。同时,根据聚焦程度测量函数设计了合理的邻域特征,完成了NSCT域高频子图的融合。实验结果表明:该算法边缘信息传递因子(Q~(AB/F))指标略低于对比算法,但空间频率(SF)、平均梯度(AG)、清晰度(SP)以及视觉信息保真度(VIFF)指标相比于对比算法分别提高了约16.31%、41.87%、19.2%以及12.07%,有效地提取了源图像更深层次的清晰测度信息,克服了稀疏表示的块效应缺陷。
In the image fusion process,the sparse representation has block processing to break the continuity of the image,resulting in serious loss of clear measurement information of the multi-focus fused image.To solve this problem,a multi-focus image fusion algorithm based on convolution sparse representation and neighborhood features is proposed.This algorithm decomposes the low-frequency subgraphs of the non-down sampling contourlet transform(NSCT) domain into the base layer and the detail layer by Gaussian filtering,and then uses the alternating direction multiplier algorithm(ADMM) to solve the sparse coefficients to complete the fusion of feature response coefficients of the detail layer.At the same time,a reasonable neighborhood feature is designed based on the focus degree measurement function to complete the fusion of high-frequency subgraphs in the NSCT domain.The experimental results show that the edge information transfer factor(Q~(AB/F)) of this algorithm is only slightly lower than that of the comparison algorithm,but the spatial frequency(SF),average gradient(AG),sharpness(SP) and visual information fidelity(VIFF) increase by about 16.31%,41.87%,19.2%,and 12.07%,respectively,compared with the comparison algorithm.The proposed algorithm effectively extracts the deeper clear measurement information of the source image,also overcomes the blockiness defect of sparse representation,and has better fusion performance.
In the image fusion process,the sparse representation has block processing to break the continuity of the image,resulting in serious loss of clear measurement information of the multi-focus fused image.To solve this problem,a multi-focus image fusion algorithm based on convolution sparse representation and neighborhood features is proposed.This algorithm decomposes the low-frequency subgraphs of the non-down sampling contourlet transform(NSCT) domain into the base layer and the detail layer by Gaussian filtering,and then uses the alternating direction multiplier algorithm(ADMM) to solve the sparse coefficients to complete the fusion of feature response coefficients of the detail layer.At the same time,a reasonable neighborhood feature is designed based on the focus degree measurement function to complete the fusion of high-frequency subgraphs in the NSCT domain.The experimental results show that the edge information transfer factor(Q~(AB/F)) of this algorithm is only slightly lower than that of the comparison algorithm,but the spatial frequency(SF),average gradient(AG),sharpness(SP) and visual information fidelity(VIFF) increase by about 16.31%,41.87%,19.2%,and 12.07%,respectively,compared with the comparison algorithm.The proposed algorithm effectively extracts the deeper clear measurement information of the source image,also overcomes the blockiness defect of sparse representation,and has better fusion performance.
引文
[1] FENG Xin,ZHANG Jian-hua,HU Kai-qun,et al.An imp-roved method for noisy images fusion[J].Journal of Optoelectronics·Laser,2017,11(28):1260-1266.冯鑫,张建华,胡开群,等.一种改进含噪声多聚焦图像融合方法[J].光电子·激光,2017,11(28):1260-1266.
[2] WANG Lu,WANG Xiao-chun.Multi-focus image fusionmethod combining W transform and NSCT[J].Journal of Electronic Measurement and Instrument,2017,31(5):756-765.王璐,王小春.W变换和NSCT相结合的多聚焦图像融合方法[J].电子测量与仪器学报,2017,31(5):756-765.
[3] WEI Li-sheng,ZHANG Ping-gai.Multi-focus image fusion method based on layered model and local restoration[J].Journal of Instrumental Instrument,2016,37(1):161-169.魏利胜,张平改.基于分层模型与局部复原的多聚焦图像融合方法[J].仪器仪表学报,2016,37(1):161-169.
[4] FENG Xin,WANG Xiao-ming,DANG Jian-wu,et al.An improved fusion method for multi-focus images based on Ripplet transform[J].Journal of Optoelectronics·Laser,2013,6(24):1204-1210.冯鑫,王晓明,党建武,等.一种改进的基于Ripplet变换的多聚焦图像融合方法[J].光电子·激光,2013,6(24):1204-1210.
[5] Yang H,Jin X,Zhou D.Block medical image fusion based on adaptive PCNN[C].IEEE International Conference on Software Engineering and Service Science,2015,330-333.
[6] CHANG Li-hong,FENG Xiang-chu,ZHANG Rui.Image fusion of quaternion wavelet transform combined with sparse representation[J].Systems Engineering and Electronics,2017,39(7):1633-1639.常莉红,冯象初,张瑞.四元数小波变换联合稀疏表示的图像融合[J].系统工程与电子技术,2017,39(7):1633-1639.
[7] SHOU Zhao-yu,HU Rong,OU Yang-ning.Improved multi-focus image fusion based on sparse representation[J].TV Technology,2014,38(7):13-16.首照宇,胡蓉,欧阳宁.改进的基于稀疏表示的多聚焦图像融合[J].电视技术,2014,38(7):13-16.
[8] YIN Wen,LI Yuan-xiang,ZHOU Ze-ming,et al.Remote sensing image fusion based on sparse representation [J].Optics Journal,2013,33(4):259-266.尹雯,李元祥,周则明,等.基于稀疏表示的遥感图像融合方法[J].光学学报,2013,33(4):259-266.
[9] Ibrahim R,Alirezaie J,Babyn P.Pixel level jointed sparse representation with RPCA image fusion algorithm[C].International Conference on Telecommunications and Signal Processing,2015,592-595.
[10] Zhao C H,Guo Y B.A fast image fusion algorithm based on sparse representation and nonsubsampled contourlet transform[J].Journal of Electronics and Information Technology,2016,38(7):1773-1780.赵春晖,郭蕴霆.一种快速的基于稀疏表示和非下采样轮廓波变换的图像融合算法[J].电子与信息学报,2016,38(7):1773-1780.
[11] OU Yang-ning,ZHENG Xue-ying,YUAN Hua.Multi-focus image fusion based on NSCT and sparse representation[J].Computer Engineering and Design,2017,38(1):177-182.欧阳宁,郑雪英,袁华.基于NSCT和稀疏表示的多聚焦图像融合[J].计算机工程与设计,2017,38(1):177-182.
[12] ZHOU Jing-hong,ZHOU Wei,ZHU Jian-jun,et al.Super-resolution reconstruction based on non-subsampled contourlet transform remote sensing image[J].Acta Optica Sinica,2015,35(1):98-106.周靖鸿,周璀,朱建军,等.基于非下采样轮廓波变换遥感影像超分辨重建方法[J].光学学报,2015,35(1):98-106.
[13] Liu X H,Chen Z B.Infrared and visible image fusion based on multiscale directional guided filtering and convolution sparse representation[J].Acta Optica Sinica,2017,37(11):103-112.刘先红,陈志斌.基于多尺度方向引导滤波和卷积稀疏表示的红外与可见光图像融合[J].光学学报,2017,37(11):103-112.
[14] Wohlberg B.Efficient algorithms for convolutional sparse representations[J].IEEE Transactions on Image Processing,2015,25(1):301-315.
[15] Tian X H,Xing W.Study of medical image fusion based on NSCT transform[J].Computer Applications and Software,2013,30(4):287-289.田秀华,兴旺.基于NSCT变换的医学图像融合研究[J].计算机应用与软件,2013,30(4):287-289.
[16] ZHANG Shen-wei,LI Wei,ZHAO Xue-jing.A fusion method of visible and infrared images based on sparse representation[J].Electrooptics and Control,2017,24(6):47-52.张生伟,李伟,赵雪景.一种基于稀疏表示的可见光与红外图像融合方法[J].电光与控制,2017,24(6):47-52.
[17] MA Miao,WAN Ren-yuan,YIN Yi-long.Multi-focus image fusion based on similarity gray relation Curvelet domain[J].Chinese Journal of Electronics,2012,40(10):1984-1988.马苗,万仁远,尹义龙.基于相似性灰关联Curvelet域多聚焦图像融合[J].电子学报,2012,40(10):1984-1988.
[18] PENG Zhen-ming,JING Liang,HE Yan-min,et al.Multi-focus image super-resolution fusion based on multi-scale sparse dictionaries[J].Optics and Precision Engineering,2014,22(1):169-176.彭真明,景亮,何艳敏,等.基于多尺度稀疏字典的多聚焦图像超分辨融合[J].光学精密工程,2014,22(1):169-176.
[19] Sheikh H R,Bovik A C.Image information and visual quality[J].IEEE Transactions on Image Processing,2006,15(2):430-444.
[20] Hajer O,Olfa M,Ezzeddine Z.Non-subsampled shearlet transform based MRI and PET brain image fusion using simplified pulse coupled neural network and weight local features in YIQ colour space[J].IET Image Processing,2018,12(10):1873-1880.
[21] Chang H S,Zhang X P.Layer-based approach for image pair fusion[J].IEEE Transactions on Image Processing,2016,25(6):2866-2881.
[2] WANG Lu,WANG Xiao-chun.Multi-focus image fusionmethod combining W transform and NSCT[J].Journal of Electronic Measurement and Instrument,2017,31(5):756-765.王璐,王小春.W变换和NSCT相结合的多聚焦图像融合方法[J].电子测量与仪器学报,2017,31(5):756-765.
[3] WEI Li-sheng,ZHANG Ping-gai.Multi-focus image fusion method based on layered model and local restoration[J].Journal of Instrumental Instrument,2016,37(1):161-169.魏利胜,张平改.基于分层模型与局部复原的多聚焦图像融合方法[J].仪器仪表学报,2016,37(1):161-169.
[4] FENG Xin,WANG Xiao-ming,DANG Jian-wu,et al.An improved fusion method for multi-focus images based on Ripplet transform[J].Journal of Optoelectronics·Laser,2013,6(24):1204-1210.冯鑫,王晓明,党建武,等.一种改进的基于Ripplet变换的多聚焦图像融合方法[J].光电子·激光,2013,6(24):1204-1210.
[5] Yang H,Jin X,Zhou D.Block medical image fusion based on adaptive PCNN[C].IEEE International Conference on Software Engineering and Service Science,2015,330-333.
[6] CHANG Li-hong,FENG Xiang-chu,ZHANG Rui.Image fusion of quaternion wavelet transform combined with sparse representation[J].Systems Engineering and Electronics,2017,39(7):1633-1639.常莉红,冯象初,张瑞.四元数小波变换联合稀疏表示的图像融合[J].系统工程与电子技术,2017,39(7):1633-1639.
[7] SHOU Zhao-yu,HU Rong,OU Yang-ning.Improved multi-focus image fusion based on sparse representation[J].TV Technology,2014,38(7):13-16.首照宇,胡蓉,欧阳宁.改进的基于稀疏表示的多聚焦图像融合[J].电视技术,2014,38(7):13-16.
[8] YIN Wen,LI Yuan-xiang,ZHOU Ze-ming,et al.Remote sensing image fusion based on sparse representation [J].Optics Journal,2013,33(4):259-266.尹雯,李元祥,周则明,等.基于稀疏表示的遥感图像融合方法[J].光学学报,2013,33(4):259-266.
[9] Ibrahim R,Alirezaie J,Babyn P.Pixel level jointed sparse representation with RPCA image fusion algorithm[C].International Conference on Telecommunications and Signal Processing,2015,592-595.
[10] Zhao C H,Guo Y B.A fast image fusion algorithm based on sparse representation and nonsubsampled contourlet transform[J].Journal of Electronics and Information Technology,2016,38(7):1773-1780.赵春晖,郭蕴霆.一种快速的基于稀疏表示和非下采样轮廓波变换的图像融合算法[J].电子与信息学报,2016,38(7):1773-1780.
[11] OU Yang-ning,ZHENG Xue-ying,YUAN Hua.Multi-focus image fusion based on NSCT and sparse representation[J].Computer Engineering and Design,2017,38(1):177-182.欧阳宁,郑雪英,袁华.基于NSCT和稀疏表示的多聚焦图像融合[J].计算机工程与设计,2017,38(1):177-182.
[12] ZHOU Jing-hong,ZHOU Wei,ZHU Jian-jun,et al.Super-resolution reconstruction based on non-subsampled contourlet transform remote sensing image[J].Acta Optica Sinica,2015,35(1):98-106.周靖鸿,周璀,朱建军,等.基于非下采样轮廓波变换遥感影像超分辨重建方法[J].光学学报,2015,35(1):98-106.
[13] Liu X H,Chen Z B.Infrared and visible image fusion based on multiscale directional guided filtering and convolution sparse representation[J].Acta Optica Sinica,2017,37(11):103-112.刘先红,陈志斌.基于多尺度方向引导滤波和卷积稀疏表示的红外与可见光图像融合[J].光学学报,2017,37(11):103-112.
[14] Wohlberg B.Efficient algorithms for convolutional sparse representations[J].IEEE Transactions on Image Processing,2015,25(1):301-315.
[15] Tian X H,Xing W.Study of medical image fusion based on NSCT transform[J].Computer Applications and Software,2013,30(4):287-289.田秀华,兴旺.基于NSCT变换的医学图像融合研究[J].计算机应用与软件,2013,30(4):287-289.
[16] ZHANG Shen-wei,LI Wei,ZHAO Xue-jing.A fusion method of visible and infrared images based on sparse representation[J].Electrooptics and Control,2017,24(6):47-52.张生伟,李伟,赵雪景.一种基于稀疏表示的可见光与红外图像融合方法[J].电光与控制,2017,24(6):47-52.
[17] MA Miao,WAN Ren-yuan,YIN Yi-long.Multi-focus image fusion based on similarity gray relation Curvelet domain[J].Chinese Journal of Electronics,2012,40(10):1984-1988.马苗,万仁远,尹义龙.基于相似性灰关联Curvelet域多聚焦图像融合[J].电子学报,2012,40(10):1984-1988.
[18] PENG Zhen-ming,JING Liang,HE Yan-min,et al.Multi-focus image super-resolution fusion based on multi-scale sparse dictionaries[J].Optics and Precision Engineering,2014,22(1):169-176.彭真明,景亮,何艳敏,等.基于多尺度稀疏字典的多聚焦图像超分辨融合[J].光学精密工程,2014,22(1):169-176.
[19] Sheikh H R,Bovik A C.Image information and visual quality[J].IEEE Transactions on Image Processing,2006,15(2):430-444.
[20] Hajer O,Olfa M,Ezzeddine Z.Non-subsampled shearlet transform based MRI and PET brain image fusion using simplified pulse coupled neural network and weight local features in YIQ colour space[J].IET Image Processing,2018,12(10):1873-1880.
[21] Chang H S,Zhang X P.Layer-based approach for image pair fusion[J].IEEE Transactions on Image Processing,2016,25(6):2866-2881.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |