Application Of Imaging Sonar Data Processing In Sinking And Revetment Engineering

In this paper,two sets of measurement schemes are designed for the lap distance measurement problem in the sinking and revetment project,and the lap distance measurement is realized from the perspective of 3D point cloud data processing and sonar image processing.The first step in the 3D point cloud data processing scheme is point cloud preprocessing,which aims to segment the target area from the initial point cloud and eliminate interfering point cloud.In this part,the point cloud segmentation method based on RANSAC theory is used,and the extra-point filtering method based on statistical analysis/point cloud density is compared.It is concluded that the extra-point filtering method based on statistical analysis is more suitable for the point cloud data in this paper.The second step is point cloud feature extraction,the purpose is to extract the relevant feature parameters in the target point cloud.In this part,the cluster classification method based on Euclidean distance is used to classify the target point cloud.An edge cluster extraction method is designed to extract the edge cluster containing feature information.The point cloud is fitted to the edge cluster and tested.The LMedS method and the M-estimator method are used to select appropriate fitting methods and parameters by quantitative analysis of the fitting results.The measurement of the lap distance is completed using the extracted characteristic parameters.In this paper,the sonar image processing scheme is designed.The steps are: image preprocessing,image feature extraction,and lap distance measurement.In the preprocessing part,an improved nonlinear filtering is proposed,which can enhance the image while maintaining the edge of the image,combined with image hole filling and open/close operations,and enhances the image recognizability.In the feature extraction part,Canny edge extraction and Hough transform line detection are used to extract the feature line from the image edge line.The arrangement overlap distance is measured by the relationship between the characteristic lines.The error analysis of the measurement results of the two sets of schemes is carried out,and the conclusion is drawn that the measurement results of the two sets of schemes are equivalent,and both can achieve low error measurement.The relationship between the two sets of measurement schemes is analyzed and the correspondence of the characteristic parameters is established.