The Design Of Marine-Acoustic-Tomography Database And Visualization Software System

Marine data is characterized by large scale,complex structure,heterogeneity from the same source,and strong spatiotemporal dependence.Therefore,it is the cornerstone to build a data management system,which differs from different characteristic of marine data,to step forward to a variety of oceanographic researches.Marine visualization is the expression of both data parameters and results from data processing.At present,the research of the management and visualization of marine data mainly focuses on a specific kind of data or an algorithm,hence bringing difficulty to apply to large-scale,multi-type and complex marine data.Therefore,based on the requirements of the project,this thesis studies three key technologies of data management and visualization for Marine data,including data management technology,image superresolution reconstruction and 3D visualization algorithm.In the meantime,the thesis develops the Marine-Acoustic-Tomography Database and Visualization Software System.The main idea of marine data management technology includes data encoding,self-adapting storage scheme and retrieval scheme.Each code,which includes different characteristic of marine data,is a unique identification of the data.It is generated during the storing process and mainly used for locating the specific data during retrieval.Based on the characteristics of the Marine data,the structure of the corresponding database is designed.In the process of storage,the name of the database and the table where the import marine data shall be stored can be decided automatically according to the structure of the database.The Marine database is established based on the above three parts to serve as the data source for super-resolution reconstruction and Marine data visualization.Super-resolution reconstruction technology aims to raise the scale of marine data to achieve the requirements of the data visualization.In this thesis,the graph signal process with two different regularization methods will be applied to the reconstruction of marine data.With the raise of the resolution,the method has shown better reconstructing effect at the edge of the invalid data.The reconstruction results can be used for the subsequent realization of 3D marine data visualization.In this thesis,different methods of the visualization of 3D Marine data are studied and especially,the ray-casting algorithm.The thesis analyses the deficiencies of raycasting when it's applied to marine data and therefore,improves it by homogenization of the vertical part of the data,data classification,the judgement of the invalid data and the simplification of the resampling process.Based on the research of three key technologies,this thesis designs a set of software with database called Marine-Acoustic-Tomography Database and Visualization Software System to encode,store,retrieve and display a variety of marine data for the related project.