Research on land use classification of Hulun Lake basin based on Hyperion hyperspectral data

摘要

The hyperspectral remote sensing realizes image formation of spatial feature of the target by taking nanoscale super-high spectral resolution and meanwhile performs continuous spectral covering for dozens of and even hundreds of narrow wave bands formed on the wider wavelength range of each image element, which is with the ability of distinguishing the subtle difference of ground object spectrum, thus providing possibility for the classification of hyperspectral image of a region. However, the hyperspectral remote sensing data has a series of problems like large number of wave band, high correlation, lower spatial resolution and so on. Through the object-oriented and random forests classification methods, the utilization potentiality is designed to excavate the hyperspectral data, so as to provide help for the further research on the hyperspectral image classification. Meanwhile, it can reduce the influence of hyperspectral data defect, give play to the advantage of spectrum, and finally achieve the goal of further improving the land use classification accuracy. Taking Hyperion hyperspectral data of Hulun Lake basin as the data source in this paper, on the basis of preprocessing and dimension reduction process of data, nine kinds of typical land types (saline-alkali soil, bare land, reed, grassland with high coverage rate, grassland with low coverage rate, puddle, marsh, water body with high sand content, water body with low sand content) in the research area are selected as classification systems, and two kinds of object-oriented and random forests classification methods are adopted, so as to perform the identification and classification for ground features in the research area. And in combination with domestic No. 1 high resolution satellite image, the precision test and reason analysis are performed. Therefore, scientific basis is provided for the hyperspectral remote sensing data processing and its application in land use classification in this paper.