Ship Wake Detection In Tiangong-2 Interferometric Imaging Radar Altimeter Image

Ships play very important roles in ocean exploration and marine economy development,and there are a lot of applications not only for civilians but also for the military for effective monitoring of ships.For example,in civilian applications,it can monitor water traffic,rescue vessels in ocean disaster,and supervise illegal ships,such as those engaged in illegal fishing,smuggling,and oil pollution dumping;in military applications,various military ships need to be detected and monitored during wartime.Synthetic aperture radar(SAR)is widely used in ship detection because it can work all day long in all weather conditions,and SAR images are of high resolution.Indirect detection of wakes has many advantages over direct detection of ships: wakes usually much larger than ship targets,and thus the requirement for spatial resolution can be reduced.Besides,we can get the ship position as well as the direction and velocity information from a ship wake.We take the radar images from Tiangong-2 Interferometric Imaging Radar Altimeter(TG2-In IRA)for study,which is quite different from traditional SAR by using very small incidence angles 1? to 8?.The small incidence angle results in different imaging effects and characteristics.The major works of this thesis are as follows.1.The geometric model of Kelvin wake was established according to the theory of ship wavemaking resistance.The sea surface under certain wind speed and direction was simulated by the linear filtering method,and the sea surface and wake were superimposed to obtain the superimposed wave height.The backscatter coefficient of the superimposed wave height was calculated by using the physical optics method.By comparing the backscatter coefficients of ship speed,ship direction,wind speed,wind direction and radar incidence angle with different values,the factors affecting the imaging of the Kelvin wake and the special imaging characteristics under small incidence angles were obtained.2.By using the TG2-In IRA images,the automatic detection of peaks in Radon transform domain was studied to show the effectiveness in improving the contrast of wakes relative to the background.We proposed a wake contrast enhancement algorithm based on the top-hat transform.The effectiveness of the proposed algorithm was validated by processing real images.Through comparing the threshold values used for wake extraction in Radon transform,we can get the conclusions that after the wake contrast was improved,much more robust detection performance can be achieved,that is to say,the more accurate the result is,the wider the range of optimal thresholds is.Although refined Lee filtering can reduce speckle noise,it basically does not help for wake detection.3.We investigated the method for retrieving the information about the ship position,heading direction,as well as the velocity.The methods for obtaining ship velocities from narrow V wake,internal wave wake and Kelvin wake were studied.Especially,we focused on a seldom attended Kelvin-Mach wake model,and based on this model,the coupling relationship between the half angle of the Kelvin-Mach wake and the ship length and velocity was explored.