Research On The Method Of DEM Extraction Based On IKONOS Stereo Imagery

The high resolution remote sensing image is the most effective means of spatial data capture, it possesses the advantage that the others do not have in the approaches of extracting DEM. Centering on the methods about DEM generated from IKONOS stereo imagery, the paper processes a thoroughly theoretical analysis and a large quantity of experiments, the main works are as follows:Designs a pyramid delamination correlation matching scheme based on gray information.The experiment shows the scheme not only provides preferable initial value, but also improves its accuracy of image matching.Makes systemic investigation in the positioning theory of IKONOS satellite stereo imagery based on the rational function model, discusses the object space scheme and image space scheme to compensate the error of the rational function model, obtains the conclusion of promoting the positioning precision effectively only by using one ground control point.Discusses some methods to generate DEM and eliminate gross error cluster. aiming at the gross error cluster of image automatic matching dataset, adopts the detection algorithm based on the point wise, and presents corresponding alteration algorithm, the experiment proves , comparing with the detection algorithm based on the point wise, the alteration algorithm gains the same effect eliminating gross errors and its form simply and calculation easily.Writes all programs using the MATLAB language, realizes the function of DEM automatic extraction.