Design Of Grid Operators Under Regular Hexagonal Field Model

With the advent of the era of geographic big data,spatial analysis of large data volume and large scale attracts much attention.While the spatial analysis method under the field model emphasizes the continuity and integrity of space,which is very suitable for the analysis and research in a large space.Field model is an important conceptual model for people's spatial cognition,and its digitized expression is raster data model which has been widely used in GIS spatial analysis.For spatial analysis of large amount of data,it is often necessary to divide the research area through regular grids,among which the most common division unit is regular quadrilateral grid.However,it is still criticized by researchers in the fields of computer vision and pattern analysis for its inconsistency in adjacent units.But regular hexagonal grid is isotropic,it has a consistent distance and connection of nearest cells and a higher sampling rate,so it's potential in neighborhood analysis and remote sensing image processing.But,at present,there are few studies on the theory and algorithm of regular hexagonal grid,and there is a lack of data organization,basic analysis algorithm and suggested tools.Based on the existing study of hexagonal grid,this paper proposes hexagonal field model and points out the fundamental problems and relevant solutions faced by establishing a hexagon model,including: grid coordinate,namely organizing and indexing grid via the cube on hexagonal grid coordinate and orthogonal coordinate system,and realizing the coordinate transformation and distance measurement between the two coordinate systems;rasterization and resampling,that is,since there is no directly available regular hexagonal raster data,rasterization of vector data and resampling of raster data are required to obtain regular hexagonal raster data;visualization: at present,there is no hardware device that directly displays regular hexagon and it can only be visualized by software approaches,including super pixel method and geometric figure method;HexASCII file format: solving the problem of result storage and file exchange.As well as run length coding and compression of HexASCII file data supported by Hilbert curve which makes the storage of regular hexagonal raster data more efficientIn addition,on the basis of set theory and with reference to mathematical morphology,this paper studies morphological operators and spatial analysis methods under the regular hexagonal field model,mainly including morphological operators such as dilation,erosion,opening and closing operations;buffer analysis,inter-visibility analysis,kernel density analysis,Voronoi diagram,path analysis and cartographic generalization and other spatial analysis algorithms.In addition,the algorithm is illustrated with the example of indoor path analysis and land use polygon aggregation.Through experiments and comparison with regular quadrilateral grid,we find that the results under regular hexagonal grid are more natural in shape and more efficient in calculation,which preliminarily verifies its advantages in the closest neighborhood analysis.In conclusion,the problems and solutions in the establishment of regular hexagonal field model are systematically expounded,and the theoretical basis of regular hexagonal field model is strengthened in this paper.The spatial analysis ability of GIS raster data model is improved by studying the morphological operator and spatial analysis algorithm of regular hexagonal field model.The advantages of regular hexagonal field model in neighborhood analysis are preliminarily verified through comparative experiments,and the application scope of the regular hexagon field model is expanded.However,due to limitations of ability and time,there are still some deficiencies in the research content of this paper.In terms of data organization,further studies are needed on such contents as regular hexagonal raster vectorization,pyramid hierarchical visualization,etc.In terms of analytical algorithms,existing algorithms still need to be strengthened,such as weighted Voronoi diagram,dynamic path planning,multi-terrain cartographic synthesis,etc.,and algorithms in other more fields need to be developed,such as remote sensing image processing,DEM analysis,etc.