Research On System Design And Image Reconstruction Algorithms For Interferometric Synthetic Aperture Sonar

Interferometric Synthetic Aperture Sonar(INSAS) is a kind of sonar system which provides three dimensional images of underwater scene. INSAS has high resolution independent of the target distance. In this thesis the key aspects of INSAS are discussed, such as system design, INSAS imaging algorithms, simulation and motion compensation.The thesis firstly introduces the basic principle of INSAS and analyzes the parameters effect on the system error. Then the optimal baseline length and angle was provided. The author also gives the performance analysis of an INSAS system with probable parameters. It can be used to prompt the future experiment.The simulation algorithm is an important part of INSAS research. Efficient simulation algorithm is the fundament of the experiment. The thesis improves the traditional algorithm instead of a delay-table based algorithm. It can reduce the without accuracy loss. And a complex images simulation algorithm is also provided. It can directly simulate the SAS images instead of the raw signal data. The results show that it can reflect the actual scene well. Last the author gives an additional reverse algorithm. It calculates the raw signal data from the simulated images with loss of accuracy.The sonar array is placed on the towfish so that the motion error can affect the image quality. After the introduction of traditional motion compensation algorithms, DPC and DPIA, a more efficient algorithm used for the powerful point target condition is proposed. Last, the roll error estimation is discussed.INSAS imaging algorithms contain image registration, flat terrain phase effect removal, phase filter and phase unwrapping. In this thesis an INSAS signal simulation and processing platform based MATLAB is established. Based on this platform INSAS imaging algorithms are discussed. The common image registration and flat terrain phase effect removal algorithms are analyzed with radar data. A new adaptive phase filter is proposed. It can suppress noise and keep the image details. The main phase unwrapping algorithms performance is compared. Last the SAS image experiment data is used to computer interferometry images with 2-pass mode algorithm.