Research On Multi-Pass SAR Coherent Change Detection And Route Optimization Method

Synthetic aperture radar (SAR) coherent change detection method plays an important role in the field of change detection due to its combinatorial use of amplitude and phase information. Further, multi-pass SAR coherent change detection method aims at making full use of multiple SAR images which are obtained through multiple flights in same area. This method obtains better performance comparing with the traditional SAR coherent change detection method. The range of application of this method is huge,such as forest vegetation monitoring, crop growth condition monitoring, battlefield target monitoring and so on. In addition, the corresponding multi-pass flight pattern needs higher cost of flight and requires stricter SAR image formation constraints, which makes the study of route optimization more important in the actual operation of multi-pass SAR coherent change detection.This thesis researches on multi-pass SAR coherent change detection and route optimization method. The whole thesis is divided into four parts: research background,basic theory, core algorithm, simulation and experiment. The main contents are shown below:1. Studied the mechanism of SAR change detection and discussed some key indicators associated with the detection performance. The study built a theoretical basis of the follow-up study.2. Studied the main process of SAR coherent change detection and presented the 3 factors (echo signal noise, base length, registration accuracy) that affect the coherence of the target. Proposed an image registration method that combined scene rough matching with target precise registration.3. Proposed a multi-dimensional circular symmetry Gaussian distribution based multi-pass SAR coherent change detection method. The analytical expression was derived and a self-adapting threshold decision strategy was presented. The simulations and experiments corresponding to this method were carried out afterwards.4. Studied the route optimization methods focusing on multi-pass imaging pattern.The boundary conditions of multi-pass SAR change detection were deduced and the B-spline curve smoothing method was introduced together with the simulation.The methods this thesis proposed had gone through a series of simulations and experiments,which proved the effectiveness and advantages when performing the task of change detection and route optimization.