Restraint Conditions in Estimating Endmembers

DOI Web Site Web Site 被引用文献1件 参考文献3件
  • AWADU Tsutomu
    Department of Biological and Environmental Engineering, Faculty of Agriculture, Tokyo University
  • OKI Kazuo
    Department of Biological and Environmental Engineering, Faculty of Agriculture, Tokyo University
  • OMASA Kenji
    Department of Biological and Environmental Engineering, Faculty of Agriculture, Tokyo University

この論文をさがす

抄録

The linear spectral unmixing method is a commonly accepted approach to estimation of coverages of some categories in hyperspectral imagery. If, however, the values of coverages are concentrated on roughly 50%, for example in agricultural land, it is difficult for this method by itself to estimate coverages because the observed spectra obtained in hyperspectral imagery are quite similar. Therefore a solution of this problem needs to be discovered. <br>　In this paper, a new method is suggested to solve that problem. It is found that it can be applied to agricultural land by the singular value decomposing method adding a new conditional restraint. This focuses on the low correlation between the endmember of soil and that of crop. Even if coverages of all areas are 40%-60%, Maximum error is shown to be less than 0.018, sufficiently small. Additionally, accurate estimation of coverages is proven to be possible.

収録刊行物

  • 農業気象

    農業気象 60 (5), 957-960, 2005

    日本農業気象学会

被引用文献 (1)*注記

もっと見る

参考文献 (3)*注記

もっと見る

キーワード

詳細情報 詳細情報について

問題の指摘

ページトップへ