Translator Disclaimer
8 February 2001 Pattern decomposition method for hypermultispectral satellite data analysis
Author Affiliations +
Abstract
We have developed 'Pattern Decomposition Method' (PDM) for multi-spectral satellite data based on linear mixing of three standard spectral patterns of ground objects, namely water, vegetation and soil. In this Method, the spectral reflectance of each pixel in a satellite image is decomposed into the three components and information of the spectra is represented by a set of three decomposition coefficients. The applicability of the PDM to continuous spectra of ground objects is studied in the wavelength region of 350-2500 nm. Especially for hyper-multispectral data analysis, data reduction is very important. The continuos spectral reflectance of land cover objects could be decomposed by the standard spectral patterns with accuracy of 4.5 percent. Mixing ratio of land cover objects in a pixel of satellite data could be evaluated using the linear mixing three decomposition coefficients. For detail analysis of vegetation change from vivid state to withered state, availability of a supplementary spectral pattern that rectify resonance absorption pattern of vivid standard vegetation for spectra of withered vegetation is also studied. The new vegetation index is proposed as a simple function of the pattern decomposition coefficients including the supplementary pattern. It is confirmed that RVIPD is linear to vegetation cover ratio and also to vegetation quantum efficiency.
© (2001) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Noboru Fujiwara, Akiko Ono, and Motomasa Daigo "Pattern decomposition method for hypermultispectral satellite data analysis", Proc. SPIE 4151, Hyperspectral Remote Sensing of the Land and Atmosphere, (8 February 2001); https://doi.org/10.1117/12.417005
PROCEEDINGS
11 PAGES


SHARE
Advertisement
Advertisement
Back to Top