An automatic geometric positioning method for optical satellite image assisted by SAR image

Compared to optical images, SAR image geometric modeling is simpler and more accurate, as its positioning is only related to satellite position. In order to fully utilize the advantages of SAR satellite images in geometric positioning, a high-precision automatic geometric positioning method is proposed in this paper for optical images assisted by SAR satellite images. Firstly, SAR DOMs and optical image are superimposed detection. Secondly, a multi-modal matching algorithm is used to perform image matching and control point optimization. Finally, the matched virtual control points and optical images are used as input information to perform adjustment calculations based on the Rational Function Model. Experiment results show that this method not only significantly improves the geometric positioning accuracy, but also has high computational efficiency and automation level. It has certain guiding significance for the rapid update production of topographic maps in the field of photogrammetry.