Study On Fast Geometric Correction For High Resolution Satellite Image

With the development of remote sensing technology, space sensors substantial increase in ground resolution, high-resolution satellite images to detect in the military, national defense, marine surveying and mapping, topographic mapping, surveying and mapping the sea reefs and other infrastructure, as well as other areas of surveying and mapping applications more widespread. From high-resolution satellite images to obtain geo-spatial information needs of remote sensing imagery need to be fixed in a projection system in the light of amendments to the original image and the existence of the geometric distortion (usually called the image geometric correction), image information for geometry measurement, comparison and composite analysis. Accuracy of remote sensing images for the elimination of its deformation is a key technology. Long period of time, satellite remote sensing images to correct the geometric control points are mainly based on the use of polynomial fitting method, the method needs to provide adequate distribution of the number of ground control points. However, in the desert, oceans, borders, outside the island, the western areas, due to obvious ground features, staff were unable to reach or need real-time location, ground control points are often difficult to access or even impossible. To this end, the use of high-precision satellite remote sensing data to study the lack of control points to correct the image geometry is very important. The Content and the results of this study are as follows:(1) describe the principle and method of geometric correction, contrast analysis of commonly used high-resolution satellite images of the geometric model;(2) put forward the idea of RPC-based regional network of multi-view video View video adjustment, as well as outside the parameters of single-solver, and through multi-view imaging test block adjustment system error can be eliminated to improve positioning accuracy; high-resolution images and low resolution Image adjustment can be mixed to improve the geometric precision of low-resolution images;(3) clarify the idea of the track extrapolation model, and through strict RPC model track model for the realization of the regional network adjustment; for control points without the use of the track extrapolation region model with different tracks and track extrapolation fast geometric correct positioning;Study shows that the sea islands, such as access control points in areas difficult to model based on the RPC lock adjustment technology to achieve high-precision positioning of remote sensing. At the same time, based on extrapolation of the track model can be applied to positioning control points or no points, the lack of control areas to correct the image geometry.