The Study On The Hyperspectral Database And The Spectral Matching Technique

The development of hyperspcectral remote sensing is one of the great technology breakthroughs made by human in the area of the Earth Observation. And it is also the advanced technique of the contemporary remote sensing. Since the appearance of the hyperspectral technique, it is distinctly different from the traditional remote sensing technique because the hyperspectral data has the abundant spectral information. And we can study on the image information and quantitative analysis focusing on the spectral dimension. In the hyperspectral image processing, spectral match technique is one of the key techniques to identify and classify the material in the image. In this paper, I aim to build a useful and effective database to save hyperspectral data, manage and analyze varied typical objects' spectral plots. Then, on the base of this database, develop and validate a new spectral mapping technique. As we all know, Hyperspectral imagery provides richer information. These new larger data volumes from hyperspectral sensors present a challenge for traditional techniques. For example, the identification of each ground surface pixel by its corresponding spectral signature is still difficult because of the immense volume of data. In this paper, a new approach toward mapping from imaging spectrometer data is presented, using a novel approach extracting the feature of wavelet vectors. Spectral data reduction using wavelet decomposition could be useful. This is because it preserves the distinctions among spectral signatures. Also it is due to the intrinsic properties of wavelet transforms that preserves high- and low-frequency feature, therefore preserving peaks and valleys found in typical spectral. So in this paper, we can get the imagery classification by computing the wavelet-feature distance between each pixel and some reference spectra.