Research On High Resolution Reconstruction Method Of Side Scan Sonar Image

Side scan sonar is a commonly used equipment in underwater detection.High-resolution side scan sonar image(i.e.sonogram)can provide richer information of the seabed environment and can obtain more adequate space information of underwater target.Therefore,it is great significance to study the reconstruction of high-resolution side scan sonar image.In this paper,the research consists of two parts: sonar target position correction and highresolution image reconstruction image.The research combines the reconstructed highresolution sonar image with accurate underwater target position to build an intuitive and complete high-resolution side scan sonar image.The main research contents are as follows:(1)This chapter studies the preprocessing technology of side scan sonar attitude data.The method analyzed the characteristics of three sonar gesture data,namely roll,pitch and yaw,and preprocessed the data according to gesture change rules.In this thesis,the window moving polynomial least square method is used to smooth the attitude data.The experimental results can be used for further research of position correction.(2)Combine gesture change with underwater target to correct side sweep position.This chapter describes the side-scan sonar system,the influence factors of sonogram and the types of distortion.At the same time,this thesis analyzes the specific influence of three sonar gestures on the underwater target,and proposes a target position correction method based on gesture change.The method calculates the relative position of the seabed target and the sidescan sonar in the local coordinate system by establishing multiple coordinate systems.At the same time,the gesture rotation matrix is constructed to correct the underwater target.Finally,the method unifies the coordinate system and gets the accurate coordinate of the side-scan target.This method can eliminate both the grayscale distortion caused by the gesture change of the side scan sonar and the geometric distortion caused by the target position deviation of the sonar image.And get the exact target position.(3)We studied a high-resolution sonar image reconstruction method of sonar image based on cone cell distribution.According to the structure and distribution of the retina of the human visual system,this research uses the distribution curve of retinal cone cells to construct the sonar seabed distribution model.Secondly,the method calculates the probability of each pixel in the sonar image coordinate according to the model.Finally,it takes the probability as the weight,and use the sonar echo data to determine the final gray value of the pixel.The experimental results show that the reconstruction method is intuitive and effective in terms of similarity and clarity,signal-to-noise ratio,average gradient and contrast,and improves the resolution of side-scan sonar image.