Quality Evaluation And Multi Scale Generalization Of Land Use Patch Based On Semantic Priority

Land use data is a reflection of the state of land use,distribution structure and its dynamic change of data,it is an important part of geographic data,there is no overlap,seamless,full coverage of the distribution characteristics in the space,and there are many levels of semantic structure on the attribute,it is the most important and the most intuitive reference of land resources management,planning and decision-making,however,different management departments,different decision-making needs different scale land use map as a scientific reference,the change process of map scale is the cartographic generalization.With the continuous maturity of a variety of cartographic generalization algorithms,how to carry on the quality evaluation of the result of cartographic generalization has gradually received extensive attention,in order to better and more accurate carry out land planning and decision-making,it has important significance of multi scale generalization of land use data and quality evaluation.It is different from the cartographic generalization of traditional map,in the process of cartographic generalization of land use data should not only consider the theoretical constraints of cartographic generalization,at the same time,it should also consider the semantic hiberarchy and the topology structure of land use data itself,cartographic regional geographical features,etc.to meet the comprehension and readability of the results of cartographic generalization.Therefore,on the basis of previous studies,on the premise of the semantic of land use patch,combining with its characteristics of the continuous space distribution,In Microsoft Visual Studio 2010 environment,with the help of component for secondary development of ArcGIS Engine10.0,constructing a cartographic generalization model that based on semantic priority of land use patch.It mainly contains:patch aggregation algorithm that based on the skeleton line of constrained Delaunay triangulation,patch amalgamation algorithm that based on buffer overlap,dimension reduction of long and narrow terrain based on extracting the centerline,etc.and we evaluate the results of cartographic generalization of different scales from two aspects that contain semantic information and spatial information.This paper takes the land use data of 1:10000 of Sanjianbao and Shuishiying Sub-district of Dalian Lushun Port Area in Liaoning province as the research object,We use the above algorithm to generalize it to 1:25000?1:50000 and 1:100000,compare the area of land use type and the number of patch that before and after generalization at different scale,and quantify the information before and after cartographic generalization from semantic and spatial perspectives,then make a comprehensive analysis of the generalization quality.The research shows that the method that generalization of land use patch based on semantic priority have a good effect to maintain the balance of area.The results of generalization at the scale of 1:50000 and 1:25000 could basic maintain the original geographical features of the cartographic area,for the results of generalization at the scale of 1:100000,the information of semantic and spatial lacks seriously because the large span of scale.With the decrease of generalization scale,patch aggregation increased,the change of patch's shape is big,the lack of map information is more seriously.At the same time,the terrain factors of cartographic area and the distribution state of the patch of original data is also an important cause that lead to the loss of map information,this will be the next step of research direction.