Research On The Visualization Method Of Discrete Global Grid Based On GPU Acceleration

Discrete Global Grid System(DGGS)simulates the Earth’s surface through continuous subdivision of the(ellipsoid)sphere to a certain level.The subdivision units of DGGS are nested at different levels,and the grids at different levels have strict transformation relations,so it can be used to uniformly describe and express the geospatial data with uneven distribution and different scales.With the improvement of the level,the complexity of grid visualization increases geometrically,which forms a barrier for the expression of high-resolution spatial data.Therefore,realizing bearing and rendering of high-precision data on a global scale on the basis of integrated data presentation and performance can promote the development of global discrete grid applications.This paper is based on the existing research of global discrete grid,considering the unity of the adjacency of the diamond grid and the most efficient data index,so the diamond grid is used as the starting point to study the visualization method of the global discrete grid.The global discrete grid is a hierarchically nested multi-resolution data model.This article first uses the characteristics of the grid itself to organize and store the diamond grid data,builds an organizational storage model for the diamond grid data,and then uses the global discrete grid to analyze Based on the characteristics of the grid and Graphics Processing Unit(GPU)related technology,a fast grid algorithm is proposed to accelerate the acquisition of the visualization framework.Finally,a grid visualization method is proposed on the basis of the former two to achieve the grid’s high global scale.Carrying and rendering of precision data.At the end of this article,a prototype system is built to verify the feasibility and correctness of the visualization method of this article.The research contents and achievements of this paper mainly include:(1)This article studies a method of organizing and storing diamond grid data,using the characteristics of the grid itself to realize the scientific storage of grid data.Based on this,this paper explores the organization and storage model of massive grid data to prepare data for the later visualization algorithm.(2)This paper studies a fast subdivision method of diamond grid.Grid subdivision is an essential part of grid visualization process,but it has a performance bottleneck,which cannot support the bearing and expression of high-precision data.Therefore,this paper proposes a diamond grid subdivision method based on GPU to realize the fast subdivision of high-level grid.(3)This paper studies a visualization method to achieve the purpose of the purpose of the diamond grid capable of carrying and expressing high-precision data on a global scale.Based on the organization and storage of grid data,this paper proposes a corresponding visualization method based on the organization and storage of grid data,explores the organization model and scheduling scheme of grid data in the GPU,and makes full use of the ability of GPU to accelerate the whole visualization process,it provides a new idea for the visualization method research of global discrete grid.The experimental results show that the global discrete grid data visualization method based on GPU acceleration realizes science organization and storage of diamond grid data,and realizes scientific organization and storage of massive diamond grid data;this paper realizes the fast subdivision of diamond grid at the GPU level,which overcomes the performance bottleneck problem of the current global discrete grid subdivision;Based on the previous data storage scheme,this paper proposes a visualization scheme of the diamond grid and accelerates the visualization process.