An Automatic Detection Method For Tensile Deformations In Optical Satellite Orthophoto

In order to realize automatic detection of tensile deformations in orthophoto which was made from satellite images, and improve the efficiency of orthophoto's production and inspection, an automatic detection method for tensile deformations in optical satellite orthophoto is proposed. Firstly, the size and range of orthophoto after ortho rectification were obtained by iterative solution based on original satellite images, DEM and the corrected RPC (Rational Polynomial Coefficient) parameters. Secondly, the image point coordinates on the original satellite image of each pixel on orthophoto were calculated by using the rational function model (RFM) and RPC parameters. Thirdly, the overlap times of the original satellite images corresponding to every pixel and its neighboring pixels were counted to determine whether the pixel was a tensile deformed pixel. If a pixel was deformed, it was marked with 255 on the binary image, otherwise with 0. Finally, the vector data of tensile deformations in orthophoto was obtained by image erosion, image expansion and vectorization of the binary image.