Research On SAR Imaging Simulation And Target Detection Method Of Ship Wake

High-resolution synthetic aperture radar(SAR)detection technology can achieve macro,long-term,continuous and dynamic observation of various targets in the complex marine environment because of its all-weather,all-weather,large-scale,multi-parameter and other characteristics.It is one of the most effective means of monitoring marine vessels.However,in actual situations,moving ships often have position shifts and blurs in SAR images.At the same time,in military applications,the rapid development of stealth technology has also brought certain difficulties to SAR image ship monitoring.In view of the above problems,indirect detection of ship targets through the characteristics of ship wakes has become a practical method.In this paper,a systematic study of the SAR detection process and target detection method of the wake of the surface ship is carried out.First,according to the hydrodynamic model of the Kelvin wake,the geometry of the wake wave of the simple Wigley hull at different speeds is calculated.Based on the sea spectrum model,The swell model and linear filtering method are used to model the shape of the linear sea surface,describe the nonlinear characteristics of the sea surface based on the high-order nonlinear wave theory,and superimpose the wake wave to simulate the Kelvin wake left by the ship when sailing on the sea surface form.Aiming at the problem of SAR imaging simulation of wake targets,based on the theory of composite surface scattering,this paper establishes a surface element scattering model for analyzing large-scale electromagnetic scattering characteristics of the sea surface by using the first-order perturbation model and Kirchhoff approximation model.The calculation results of the model show that the electromagnetic scattering distribution of the sea surface composite structure shows clear texture characteristics,and the distribution characteristics are affected by the radar's own parameters,the marine environment and the movement of the ship.On this basis,according to the modulation mechanism of the wave signal on the radar signal,this paper modulates the surface element scattering distribution,and then converts the SAR imaging resolution unit into three-dimensional coordinates to establish a projection mapping relationship between the image plane and the sea level,and then realizes the ship's wake Simulation imaging of target SAR detection process.Analyze the simulation results,and further obtain the influencing factors and changing laws of the wake SAR detection characteristics.Based on the calculated data and relevant observations,this paper further studies the applicability and feasibility of target detection methods in the field of deep learning for wake detection.In order to obtain sufficiently rich and high-quality training samples,this paper based on the generative adversarial network(GAN)network training and generation of the original image data,obtained a large number of new data,expanded the sample set,and then based on the expanded sample Train thetarget detection network,learn the abstract features in SAR images,and realize the detection and recognition of wake targets.The test results show that sea conditions,imaging relationships,and polarization methods all have a certain impact on the accuracy of target detection.At the same time,compared with traditional linear detection,this method can better identify the wake characteristics of high noise,weak targets,non-linear shape and other situations.