Research On Three-Dimensional Reconstruction Method Of Submarine Pipeline Based On Line Laser

In the external inspection of submarine pipelines,it is necessary to obtain accurate external three-dimensional information of the pipeline in order to determine its operating status.Traditional acoustic measurement has good stability.But its measurement resolution is low,so that it is impossible to obtain more detailed information about the pipeline.The threedimensional measurement technology based on line laser has high resolution for the measurement of short-range underwater objects,and can obtain a fine three-dimensional data of the surface of the object.Therefore,this thesis studies some key technologies of the line laser measurement system.The main research contents are as follows:First,the transmission characteristics of blue-green lasers in the marine environment are analyzed,and an underwater laser transmission model is constructed based on the principle of the line laser measurement system.The Monte Carlo method is used to simulate and analyze the characteristics of laser imaging in the marine environment to clarify the characteristics of underwater laser imaging.Secondly,in view of the fuzzy edges and low contrast of the underwater laser image,the underwater image is denoised by bilateral filtering.The gamma transform is used to enhance the contrast of the laser stripe area.The image background noise interference is suppressed,and the underwater image quality is improved.At the same time,in order to improve the accuracy of extracting the center of the laser stripe,an adaptive gray-scale gravity center method based on Gaussian weighting is proposed.It can quickly obtain accurate surface contours and improve the accuracy of the three-dimensional reconstruction.Then,the three-dimensional measurement data processing algorithm is researched.According to the position of the noise distance from the data subject,statistical filtering and an improved bilateral filtering algorithm are used to filter out the noise.It can reduce the interference of noise on the real details of the three-dimensional point cloud model,and avoid errors in the subsequent registration of the three-dimensional point cloud data.For the registration problem of point cloud models with less overlapping data under different perspectives,the point cloud feature point extraction method based on the normal vector difference feature is improved.The uniform block extraction method makes the surface feature point distribution of the point cloud model more balanced.Therefore,the feature point set is used for initial registration and point-to-plane iterative closest point registration.And error corresponding point pairs are eliminated,so that the registration accuracy and speed of the point cloud model are improved.Finally,by setting up a line laser scanning system to conduct underwater pipeline measurement experiments in a pool,accurate three-dimensional point cloud data on the outer surface of the pipeline are obtained,which proves that the underwater line laser scanning system based on the algorithm of this thesis can realize the rapid and high-precision measurement of the submarine pipeline surface.