Research On Moving Target Detection Methods For Video SAR Based On Shadows

Video SAR is a novel radar technology that not only possesses the advantages of traditional SAR,such as all-weather and all-time imaging,but also has an extremely high frame rate and resolution.This allows for real-time,high-precision imaging of complex environments,providing superior services for military and civilian applications.Moving target detection and localization is an important research topic in the field of video SAR,with immense research value and prospects.This thesis mainly focuses on the study of moving target shadow detection methods,with the main research content as follows:1.Focusing on video SAR imaging under circular beamforming mode,we analyze the similar shadow formation mechanism of static and moving targets in the radar imaging process.Through simulation experiments,we demonstrate the relationship between the shadow quality,the speed of the moving target,and the synthetic aperture time.We also briefly introduce the basic concept and common equations of dynamic programming,and provide an analysis of the motion model and measurement model of the DetectionBefore-Tracking algorithm(DP-TBD)based on dynamic programming.2.To address the problem of moving target detection in video SAR imaging results,a detection-preceded tracking method based on dynamic programming is proposed.This method utilizes the continuity of moving target trajectories in video SAR images,accumulating shadow target amplitude through multiple frames,distinguishing it from background noise,and completing the detection of moving target shadows.Detection experiments using Sandia National Laboratories(SNL)video SAR data verify the performance of the proposed method.To further reduce detection errors,factors influencing shadow detection are analyzed.Based on this method,an improved approach is proposed,which employs velocity vectors to weight intermediate value functions and optimizes the transition state set.The detection performance of this method is verified through simulation experiments,and a comparison of the detection results of the two methods is provided.3.To address the problem of varying target numbers in video SAR imaging area scenes,a sliding window-based DP-TBD algorithm is proposed.Multiple sliding windows are used to detect targets with varying numbers,and the effectiveness and detection performance of the method are verified through simulation and SNL video data.Based on this,a region-based shadow detection method is proposed.This method,while maintaining the same detection performance,has a lower computational complexity and can more efficiently complete shadow detection and target tracking.The detection performance of this method is verified through simulation experiments and SNL released video SAR data detection experiments.