Study On Geometric Precision Correction Of Space-borne SAR

Nineteen fifties,American Wiley(Carl Wiley)proposed that it could improve the angular resolution of the radar by the means of frequency analysis method.Then,University of Illinois of the United States confirmed the principle of synthetic aperture and the prototype of Synthetic Aperture Radar is formed.Synthetic Aperture Radar independent emission of electromagnetic waves and receives the echo reflected or scattered by the object,so it can be in any weather environment,any time of the imaged observation.However,radar side-looking and the undulated terrain cause the image to appear foreshortening,layover,shadow and other geometric distortions.In the early period of relatively backward measuring technology,positioning error and loss of accuracy due to the approximate hypothesis is acceptable.As the accuracy of track and measurement improved,how to effectively restrain the influence of the satellite orbit measurement error,radar system error and terrain fluctuation and other factors on the geometrical characteristics of SAR image,and improve the geometric quality of SAR image,it becomes one of the key issues of high resolution space-borne SAR image processing.In this paper,the theory and experiment of the control point extraction based on orthography,the foundation of high accuracy control point geometric precision correction,and the DEM aided correction of SAR terrain.The work and innovation of this thesis are mainly:Chapter 2 studies the extract method of control point based on orthography,the image registration method based on the SIFT algorithm is applied to orthography and SAR image.A kind of false matching point elimination method based on Scale factor abnormity detection is put forward,the effectiveness of the control point extraction method is verified by using the image registration method.Chapter 3 studies SAR image geometric correction model based on control point.Firstly,the geometric correction method of affine transformation is studied.The model parameters are not clear and the positioning accuracy is not high.Then,the geometric correction model based on system error estimation is established.The model parameters are simplified by using the SAR image positioning principle and the error transfer model.The simulation experiments show that the new model has the advantages of clear physical meaning and higher positioning accuracy.Chapter 4 researches DEM aided geometric correction method of SAR image.The influence of elevation fluctuation on SAR image positioning is analyzed.Based on this,the geometric correction method of SAR image aided by DEM and the system calibration method of DEM simulation SAR image are studied.The validity of the method is verified by simulation data and TerraSAR-X real data.Chapter 5 is the conclusion and prospect.The main work in this dissertation is summarized,and the future work is discussed based on the research of current development and the presenting demands of the practical application.