• DocumentCode
    2123325
  • Title

    Physical investigation on biochemical prediction using continuum removal

  • Author

    Shi, Runhe ; Zhuang, Dafang ; Niu, Zheng

  • Author_Institution
    Inst. of Geogr. Sci. & Natural Resources Res., Chinese Acad. of Sci., Beijing, China
  • Volume
    2
  • fYear
    2004
  • fDate
    20-24 Sept. 2004
  • Firstpage
    1463
  • Abstract
    This paper examines physical bases for biochemical prediction with hyperspectral data. A distinctive absorption feature centered at 2100 nm in dry leaves is focused on. Continuum removal is employed to the original reflectance spectrum from 2030 nm to 2220 nm to eliminate the influence of background. After comparing some representative samples selected from NASA ACCP data set, two features due to different carbon and nitrogen concentrations are discovered in the continuum-removed spectra. (1) The continuum-removed spectrum curve from 2030 nm to 2050 nm are nearly straight, of which the slope is highly correlated with the carbon concentration. (2) An inflexion located at 2054 nm indicates the nitrogen concentration, which is caused by an absorption pit of protein. The slope change is obvious when the nitrogen concentration is high. Both of the features are caused by essential and unique absorption features of cellulose, lignin and protein in this region, which make them physically based and widely applicable.
  • Keywords
    carbon; nitrogen; remote sensing; spectral analysis; 2030 to 2220 nm; C; N; NASA ACCP data set; biochemical prediction; carbon/nitrogen concentration; cellulose/lignin/protein absorption feature; continuum removed spectrum curve; dry leaf; hyperspectral data; physical investigation; reflectance spectrum; slope change; Absorption; Biosensors; Hyperspectral imaging; Hyperspectral sensors; Infrared detectors; NASA; Nitrogen; Proteins; Reflectivity; Remote sensing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004 IEEE International
  • Print_ISBN
    0-7803-8742-2
  • Type

    conf

  • DOI
    10.1109/IGARSS.2004.1368696
  • Filename
    1368696