Researches On The Matching Method For The Multi-source Planar Residential Area

With the continuous acceleration of China's urbanization process,urban development becomes more and more rapid.The data of urban area update cycle is shortened,so the key point is how to get the latest and high precision incremental information quickly.Now,there are two ways to obtain resident incremental information,they are ground measurement based on GPS equipment and elements extraction based on high-resolution remote sensing image.The first method requires professional measuring instruments and personnel,and this method has the disadvantages of long cycle,heavy workload and high cost.The second method is limited by the image processing technology and the complexity of urban status quo,difficult to automate,and the extraction accuracy and efficiency also need to be improved.In recent years,thanks to OSM can provide rich content,strong current situation,free access to global resident data,Research on the matching between OSM and professional mapping data can solve the problems such as the renewal of planar residential land elements.Thus,it has become one of the hot issues of surface element matching.Most of the matching methods of planar residents focus on professional surveying data.However,considering the characteristics of crowd-source data,Whether the existing methods can satisfy the matching between the crowd-source data and professional data,which has become an urgent problem to be solved in current relevant researches.In this paper,OSM residential area is the main research object,to study the matching method for the crowd sourcing surface residential area.The main work of this paper is reflected in the following aspects:(1)Conventional geometric descriptive factors of planar residents cannot properly describe the geometric features of elements,resulting in a matching failure.In this paper,we explore a planar resident land matching method based on geometric descriptors in polar coordinates.This method describes the geometric features of the elements in polar coordinate system,and combining with the idea of hierarchical matching.What's more,it selects the centroid factor,area overlap factor and geometric description factor in polar coordinate system.(2)The selection of candidate matching sets are too large to complete the matching process.In this paper,the spatial division method of road mesh and spatial clustering is introduced.An improved matching method of road mesh and spatial clustering is proposed.This method introduces road mesh and spatial clustering to divide the whole data set into different levels,and steps down the candidate matching set.Finally,three factors,area size,direction of elements and center of mass,are selected to construct the comprehensive index function,finally finish the matching process.(3)In order to verify the two methods in this paper,hierarchical matching method and principal component analysis method are selected as the comparison algorithm.The two comparison algorithms are tested together with the two algorithms in this paper.In this paper,OSM data and professional data are selected as experimental data for experiments.The results show that: Compared with hierarchical matching method and principal component analysis method,the results of the two improved algorithms in this paper have high recall and precision.The advantages of the improved algorithm are verified.