孙雪健等:Enhancement of Spectral Resolution for Remotely Sensed Multispectral Image
被阅读 1517 次
2015-10-23
Enhancement of Spectral Resolution for Remotely Sensed Multispectral Image
作者:Sun, XJ (Sun, Xuejian)[ 1 ] ; Zhang, LF (Zhang, Lifu)[ 1 ] ; Yang, H (Yang, Hang)[ 1 ] ; Wu, TX (Wu, Taixia)[ 1 ] ; Cen, Y (Cen, Yi)[ 1 ] ; Guo, Y (Guo, Yi)[ 2 ]
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
卷: 8  期: 5  页: 2198-2211
DOI: 10.1109/JSTARS.2014.2356512
出版年: MAY 2015
 
摘要
Hyperspectral (HS) remote sensing has an important role in a wide variety of fields. However, its rapid progress has been constrained due to the narrow swath of HS images. This paper proposes a spectral resolution enhancement method (SREM) for remotely sensed multispectral (MS) image, to generate wide swath HS images using auxiliary multi/hyper-spectral data. Firstly, a set number of spectra of different materials are extracted from both the MS and HS data. Secondly, the approach makes use of the linear relationships between multi and hyper-spectra of specific materials to generate a set of transformation matrices. Then, a spectral angle weighted minimum distance (SAWMD) matching method is used to select a suitable matrix to create HS vectors from the original MS image, pixel by pixel. The final result image data has the same spectral resolution as the original HS data that used and the spatial resolution and swath were also the same as for the original MS data. The derived transformation matrices can also be used to generate multitemporal HS data from MS data for different periods. The approach was tested with three image datasets, and the spectra-enhanced and real HS data were compared by visual interpretation, statistical analysis, and classification to evaluate the performance. The experimental results demonstrated that SREM produces good image data, which will not only greatly improve the range of applications for HS data but also encourage more utilization of MS data.
 
通讯作者地址: Yang, H (通讯作者)
Chinese Acad Sci, Inst Remote Sensing & Digital Earth, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China.
地址:
[ 1 ] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China
[ 2 ] Commonwealth Sci & Ind Org CSIRO, Digital Product & Serv Flagship, N Ryde, NSW 2113, Australia