Application Of Geoeye-1 On Investigation Of Geological Hazard In Mining Area

China is a major mining country. For many reasons, the project activities lead to a large number of geological disasters, which have repeatedly caused great losses and threats to lives and property. How to use the latest methods to improve the basic technical problems in the mining process of the remote sensing survey of geological hazards, is a strong support for geological hazards investigation.Analyzed on the basis of previous studies, the paper systematically carried out a research on remote sensing image processing and information extraction methods, and put forward new technical methods as the followings:1, For the difficulty of the control point selection in this study area with its own topographic characteristics, the paper studied DEM of the mining area deeply.By reference to Geoeye-1 data imaging time and the solar elevation angle, the topography of the test area was simulated, and simulated DOM was generated too by color film. DOM was taken as the reference image selection control points, in order to achieve correction results with high accuracy.2, Based on the existing conventional filtering algorithm, the paper introduced the concept of "peripheral field". The filter can identify the image of the "internal" and "edge" before it calculated, so as to achieve the effect that "internal" and "edge" smooth and sharpen respectively.The conflicts of edge enhancement and smoothing in conventional algorithm were initially balanced. Filtering got some improvements.3,Geoeye-1 data with its own characteristics, the thesis analyzed the PCA and the wavelet fusion comparativly with their complementary fusion strengths. These two methods are integrated through testing.There formed a scheme with the PCA fusion as basic framework, wavelet fusion program supporting the integration of enhanced spectrum. Fusion not only maintained the original resolution on the basis of PCA fusion, but also increased the amount of information to some extent.4,In the conventional object-oriented classification process, classification and decision trees are combined by GIS overlay analysis technology. The "non-isolated region of interest","Construction of multiple classification space" and "object-oriented final classification"were realized progressively in the classification process. It made the classification process more clear, the base of object extraction more full,and improved the accuracy of automatic extraction of information.In the above analysis and theoretical approach, relyed on the western hills in Beijing selected test area, the thesis chose Geoeye-1 data, did image processing, carried out geological hazard information extraction test, and achieved better accuracy. To some extent, this method is proved to be feasible.