Hyperspectral data transformation and vegetation index performance based on the universal pattern decomposition method

Traditional vegetation indices are based on only a few spectral bands. However, hyperspectral spectrometers, such as the airborne visible infrared imaging spectrometer (AVIRIS), collect data with 224 contiguous spectral bands. Traditional vegetation index extraction methods lose much of the information contained in hyperspectral data. The universal pattern decomposition method (UPDM) is tailored for hyperspectral data analysis. In this article, we consider the UPDM as a type of multivariate analysis; standard patterns are interpreted as an oblique coordinate system and coefficients are thought of as the coordinates of a pixel';s reflectance. This article describes UPDM hyperspectral data transformation of AVIRIS data, the performance of a vegetation index based on the universal pattern decomposition method (VIUPD), and the influences of a noise-to-vegetation index. The results demonstrate that the VIUPD is an effective vegetation information extraction approach for hyperspectral data. The VIUPD is more sensitive to vegetation conditions than the normalized difference vegetation index and enhanced vegetation index. Furthermore, noise influences can be neglected in VIUPD computations, with satisfactory accuracy. (c) 2007 Society for Imaging Science and Technology.