{"title":"Accuracy Analysis of Interpolation Method for Abrupt Change of Seabed Water Depth","authors":"Huadong Ma, Haiying Zhou, Yubo Wen, P. Wang","doi":"10.1109/ICGMRS55602.2022.9849371","DOIUrl":null,"url":null,"abstract":"In view of the uneven distribution of single beam measurement data density in the area with abrupt seabed terrain, it is difficult to form a refined digital water depth model to provide services for ship navigation, marine engineering construction, military activities and so on. In this paper, the area with large fluctuation of seabed topography is selected as a demonstration, and the three-dimensional interpolation methods such as linear interpolation triangulation method, nearest neighbor interpolation method and natural neighbor interpolation method are used to interpolate the single beam bathymetric data in the study area. Based on the two evaluation indexes of residual sum of squares (RSS) and root mean square errors (RMSE) and the cross validation evaluation method, the accuracy of the results of the three interpolation methods is evaluated. The final results show that when water depth interpolation is carried out in the area with large changes in seabed topography, the linear interpolation triangulation method has faster interpolation speed and the natural adjacent point interpolation method has higher interpolation accuracy. The research in this paper has practical reference value for the three-dimensional interpolation of single beam water depth data in the area with large seabed terrain changes.","PeriodicalId":129909,"journal":{"name":"2022 3rd International Conference on Geology, Mapping and Remote Sensing (ICGMRS)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 3rd International Conference on Geology, Mapping and Remote Sensing (ICGMRS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICGMRS55602.2022.9849371","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In view of the uneven distribution of single beam measurement data density in the area with abrupt seabed terrain, it is difficult to form a refined digital water depth model to provide services for ship navigation, marine engineering construction, military activities and so on. In this paper, the area with large fluctuation of seabed topography is selected as a demonstration, and the three-dimensional interpolation methods such as linear interpolation triangulation method, nearest neighbor interpolation method and natural neighbor interpolation method are used to interpolate the single beam bathymetric data in the study area. Based on the two evaluation indexes of residual sum of squares (RSS) and root mean square errors (RMSE) and the cross validation evaluation method, the accuracy of the results of the three interpolation methods is evaluated. The final results show that when water depth interpolation is carried out in the area with large changes in seabed topography, the linear interpolation triangulation method has faster interpolation speed and the natural adjacent point interpolation method has higher interpolation accuracy. The research in this paper has practical reference value for the three-dimensional interpolation of single beam water depth data in the area with large seabed terrain changes.