Research On The Construction Method Of Seabed Terrain Dataset Based On The Fusion Of Multisource Information

Seabed terrain data is the basic data in many researches,such as three-dimensional display of terrain and landform,underwater navigation of submersibles,terrain matching navigation and so on.In order to solve the problems of seabed terrain data in terms of noise,abnormal values,resolution,information volume,timeliness,and data storage query,this topic focuses on seabed terrain data and breaks through many problems that affect the quality of dataset in traditional data processing methods.Improve the data quality of seabed terrain data,organize the terrain data reasonably,design efficient indexes,and realize the effective storage of terrain data and rapid processing and query.The main research work includes:Firstly,the existing seabed terrain datasets have low accuracy and resolution,less information,poor timeliness and other quality problems,which limit their application in the high-precision field.At the same time,the effectiveness of storage and the speed of querying massive terrain data need to be improved.Aiming at the drawbacks and shortcomings of the seabed terrain dataset in terms of data quality and data storage and data query,this paper proposes an overall construction plan for the seabed terrain dataset based on the demand analysis of the seabed terrain dataset.Secondly,aiming at the deficiency of the traditional bathymetric data fusion algorithm in fusing high-precision data,the optimization and improvement of traditional fusion algorithms are studied.Based on the DEM(Digital Elevation Model),the electronic chart bathymetric data and multi-beam bathymetric data are integrated to construct a new seabed terrain dataset.The electronic chart is encapsulated by S57 standard,and the encoding has strong compressibility.Therefore,the analysis of electronic chart and the extraction method of water depth data are studied.The experimental results show that the improved algorithm can combine the advantages of different data sources and improve the precision and resolution of data sets.Thirdly,aiming at the problem of the lack of multibeam sounding data in the seabed area,the amount of information on the seabed terrain data is small,this paper studies the terrain generation algorithm based on convolutional neural network style transfer.The style features of the land terrain and the content features of the seabed terrain are extracted for style transfer,and the parameters are adjusted for multiple groups of tests.The results show that the generated seabed topography not only retains the undulating form of the original seabed topography,but also contains the topographic details of the land topography,which can increase the information content of the data set and improve the timeliness of the data set.Finally,the distributed storage,spatial index structure and multi-resolution topographic pyramid model are studied to solve the problem of real-time dynamic 3d display of submarine topography.In the multi-resolution pyramid model,the space complexity and time complexity of different storage schemes are analyzed,so that technicians can choose the reasonable construction scheme according to the actual situation.The correctness test and performance test of the database show that the database can realize the effective storage of data with high accuracy.At the same time,the distributed database is suitable for processing massive data.With the increase of data volume,the advantage will be more obvious.