Research And Application Of Vectorization Algorithm Based On Remote Sensing Classification

Remote sensing data is one of the important source of information in GIS system and remote sensing applications. Remote sensing classification map is a special type of the commonly used remote sensing data. In order to keep the efficiency and practicality of the system, GIS has high requirements for the efficiency of vector algorithms and topological structural integrity. Remote sensing classification map is an important integration method of RS and GIS. Due to its many shortcomings, raster data structure, to a certain extent, can not meet the needs of GIS; however, vector data is precisely to make up for these shortcomings.In addition, the integrity of topology reconstruction and high efficient vector-raster conversion methods can also provide a better way to ensure the combination of GIS and RS. Therefore, the vectorization of remote sensing classification has great significance in the Integration of GIS and RS.In recent years, several new methods have appeared to expand the idea of solving such problems. Raster and Vector method has been greatly improved, but there are still some imperfect places. In order to enhance the integrity of topology reconstruction, this article has studied and improved the vectorization of the remote sensing classification on the basis of the previous algorithm. It has put forward a kind of vectorization algorithm based on remote sensing classification. The algorithm can improve the integrity of polygon topology in the vector and it is applied to droughts and flood and wind disaster prevention decision support system.The main contents are as follows:(1) It analyzes the core problem of remote sensing classification map vectorizaton, namely, the classification map and complex polygons, and studies the complex polygons based on topological relations. In the paper, it proposes the vectorization to resolve the complex polygons in a better way, especially, to resolve those which include the sharing-boundary islands. The sharing-boundary islands can better deal with the remote sensing map vectorization algorithm, which further improve topological structural integrity of the sharing-boundary islands vectorization.(2) When polygon is generated by the nodes and the boundary points, the first two points on arc segment determine the attribute information of its adjacent polygons and the establishment of an index on the arc. As a result, polygon is generated directly through the index rather than counting the minimum interior angle, which in some sense, increases the speed of the generated polygon.(3) The remote sensing classification map vectorizaton algorithm is applied to the droughts and wind disaster prevention decision support system. It provides the basis for spatial analysis based on vectorization such as buffer analysis, overlay analysis, network analysis, etc.