The Research On Match Method Of Multi-scale Geographic Entity Based On Grid Index And Geometric Features

Scale is an important characteristic of the geographical spatial entity, as the expression of multi-scale geographic data, the status of geographical entity cross scale matching is becoming more obvious. On one hand, through the evolution of the geographic entities geometry, form the related indexes classification system attends to spatial position and geometric features for geographical entity. This paper discusses dynamic selected and quantitative methods of the related indexes, constructing comprehensive measure model of geometry difference, and realizes the multi-scale match method system for geographical entity, which will lay a solid foundation for the spatial relationship and semantic association of geographical entity. On the other hand, using this method to train the foundation geographical samples data can further enrich the multi-scale and spatial similarity measure theory and method, at the same time provide theory and method for change detection, updating, and retrieval of geographical entity.This paper using grid index and geometric features for overall design ideas to realize multi-scale match of geographical entity; focus on two keys:grid index spatial position, similarity index selection of geometric feature matching and matching model building. The research content are as follows:1. Space position and geometric characteristics variation analysis of geographical entity under multi-scale. Position and geometric variation analysis of the multi-scale geographical entity under the condition of merger, reduction, thin, and shift, summarized variation rules and characteristics between different entities category and different polygon geometric.2. Spatial position target construction and quantify of grid index; similarity matching target quantify and selection of Geometry characteristic.3. Geometry differences comprehensive measure model building. Explore the fusion of different similarity methods and characteristics target coupling mechanism, studied the difference calculation method of characteristic target, construct a geometric difference comprehensive measure model, and then entrust weight coefficient for all the components difference, finally obtained the geometric similar degree of related entities.Conclusions have been drawn through the analysis of this paper: 1. Grid index based on entity geometrical configuration, solve local search missing the optimal matching candidate set, and remarkably improved inquires retrieval efficiency of geographic entities.2. Solve the matching failure problem of the single match method to deal with a variety of geographical entity geometry form change, make up for traditional global search efficiency and explore a new method of fast features matching.3. Provides a quick retrieval and matching method for foundation geographic information elements updated and share construction.