Study on the application of hyperspectral remote sensing in plant classification

Hyperspectal remote sensing is one of the main trends in the domain of remote sensing technology. Hyperspectral data contain plenty of information about space, radiation and spectrum, which makes plant classification more precise. In the west of China, plant distribution is heavily dispersed because the loess terrain is liable to erosion by wind or rain. This makes it very difficult to survey plant distribution using normal multispectral remote sensing methods. The paper introduces the methods of plant classification using imaging spectral data obtained by OMIS I in detail, including traditional methods after the best features selecting from hyperspectral data, and ones based on spectrum matching technique uniquely applied in hyperspectral remote sensing, such as spectral angel mapping, derivate spectrum shape matching etc. The classification result verifies the effectiveness of hyperspectral remote sensing in plant classification.