Research On 3D Reconstruction Of Side Scan Soner Image Based On SFS Method

In marine engineering,comparing with common single-beam and multi-beam systems,the side-scan sonar has the characteristics of high resolution,high efficiency and low price.It is a powerful tool in marine development,seabed terrain survey,archaeological investigation and other industries.However,the side scan sonar is affected by its measurement mechanism and operation mode,and cannot directly obtain the information of the seabed surface topography and the height of the underwater target,so the research on the three-dimensional reconstruction of the side scan sonar image is of great significance.Based on the in-depth understanding of the working principle of side scan sonar,this paper starts from the original data of sonar to study the underwater terrain 3D reconstruction system.The main research contents are as follows:Firstly,the three-dimensional reconstruction method of the side scan sonar image is researched.The thesis analyzes the characteristics of various methods of three-dimensional reconstruction of side scan sonar images,and studies the main difficulties and problems in order to pave the way for subsequent theoretical research and practice.Then decode and visualize the original data of the side scan sonar.On the basis of understanding the structure of the XTF file format,the Visual Studio platform is used to realize the acquisition of various effective information and the visualization of the two-dimensional waterfall diagram of the side scan sonar.The analysis result’s are compared with the side-scan sonar waterfall chart obtained by the foreign software SonarWaveLite to verify the accuracy of the analysis software.Secondly,the preprocessing method of the two-dimensional waterfall diagram of side scan sonar is studied.In the submarine line detection part,this paper improves on the traditional submarine line detection method,and proposes an improved submarine line detection method based on the regional growth method,which realizes the accurate detection of the submarine line.In the gray-scale correction part,an improved gain compensation method is proposed,which achieves a more balanced compensation result compared to the two traditional gain compensation methods,and uses experiments to verify its effectiveness.Finally,the research on the 3D reconstruction of the side scan sonar based on the shape from shading(SFS)method is completed.On the basis of the existing algorithm,the SFS minimization method is improved by wavelet transform,and the three-dimensional reconstruction of underwater terrain is realized.The virtual image and actual image are used as input,and the algorithm proposed in this paper is analyzed through comparative experiments.The results show that compared with the two classic SFS algorithms,the performance of the SFS minimization method based on wavelet transform is significantly improved and can be clearly displayed.The undulations of the seabed surface and the subtle features of the underwater topography.