期刊文献+

结合空间-光谱调制及图像分割的多光谱图像融合方法 被引量:4

A Pansharpening Based on Spatial-Spectral Modulation and Cooperation with Segmentation
下载PDF
导出
摘要 为了进一步提高多光谱(MS)图像与全色(PAN)图像之间的融合质量,平衡空间细节的注入与光谱信息的保持,提出了一种基于局部自适应空间-光谱调制与图像协同分割的融合方法.该方法利用k-means算法、根据MS图像的光谱特性进行图像分割,得到不同的连通体组,进而基于局部连通体组构建了局部自适应光谱调制(LASpeM)系数和局部自适应空间调制(LASpaM)系数,分别对融合图像中的光谱与空间信息进行调制;其中,LASpeM系数的构建基于MS和PAN图像中的细节提取以及MS波段之间的光谱关系,LASpaM系数的构建则基于MS和低分辨率PAN图像之间光谱特性的局部差异及相关性.另外,引入融合与分割的协同思想,利用图像分割来优化融合结果,并根据融合结果的反馈信息对分割算法的参数进行调整.在Matlab环境下,采用2个卫星GeoEye-1和QuickBird数据集进行融合实验,结果表明,文中方法在主观视觉与客观评价指标方面总体上优于7种经典及流行的融合方法,能够平衡融合图像的空间信息注入和光谱信息保持,有效地减少光谱扭曲. In order to improve the fusion quality of the multispectral(MS)and panchromatic(PAN)images and achieve a balance between the injection of spatial details and the preservation of spectral information,a pansharpening method based on local adaptive spatial-spectral modulation and cooperation with segmentation is proposed in this paper.k-means algorithm is used to segment MS images into different connected component groups according to their spectral characteristics,then the local adaptive spectral modulation(LASpeM)and local adaptive spatial modulation(LASpaM)coefficients can be acquired based on local component groups.LASpeM coefficient matrix is estimated based on details extracted from MS and PAN images and also spectral relationship between MS bands;LASpaM coefficient matrix is constructed based on local deviation and correlation between the spectral characteristics of MS and low-resolution PAN images.Moreover,cooperation with segmentation is introduced to pansharpening in this paper,the LASpeM and LASpaM coefficients are estimated based on component groups to optimize the fusion image,and the feedback from fusion result is applied to adjust the parameters of the segmentation algorithm.The evaluation of the proposed method is performed in Matlab environment based on data sets from the GeoEye-1 and Quick-Bird satellites.Experimental results show that the proposed method achieves better visual results and objective evaluation indices than seven classic and state-of-the-art fusion methods,and prove that the method is able to balance the spatial and spectral information while reducing spectral distortion.
作者 焦姣 吴玲达 王朴军 Jiao Jiao;Wu Lingda;Wang Pujun(Science and Technology on Complex Electronic System Simulation Laboratory,Space Engineering University,Beijing 101416;32026 of PLA,Kaifeng 475000)
出处 《计算机辅助设计与图形学学报》 EI CSCD 北大核心 2019年第12期2101-2112,共12页 Journal of Computer-Aided Design & Computer Graphics
基金 国家自然科学基金(61801513) 装备预研基金(61420100103)
关键词 多光谱与全色图像融合 融合与分割的协同 局部自适应光谱调制 局部自适应空间调制 K-MEANS算法 fusion of the multispectral and panchromatic images cooperation between pansharpening and segmentation local adaptive spectral modulation local adaptive spatial modulation k-means algorithm
  • 相关文献

参考文献3

二级参考文献22

共引文献27

同被引文献46

引证文献4

二级引证文献11

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部