General and specific methods studying on bands selection of hyperspectral remote sensing data

To hyperspectral remote sensing application, it is a bottleneck for its numerous bands, which causs problems of mass data. This paper discusses how to select optimal bands from hyperspectral remote sensing data, and promotes general methods and specific methody. First, the thesis discusses general methods, which are common for processing all kinds of hyperspectral data by considering the information content of images and correlation between hands and tire used as rough selection. Then it focuses on studying how to select bands for specific tasks and objects, which are called as specific methods and used as accurate selection. Specific methods consider hands combination, spectrum differences, and separability of objects.