Study On Automatic Selection Method Of Urban Road Network For Map Generalization Combined With POI Data

Map generalization is the core of map design and production,and it is also one of the basic methods in the process of spatial data scale transformation and data fusion.Map generalization contains creative thinking and is the most challenging and innovative research field in modern cartography study.As an important geographical element,roads have always been one of the main objects of map generalization.The process of road generalization includes the selection and simplification of road networks.Among them,the selection of road is a precondition for road simplification and a key for road generalization.The existing research on the automatic selection of roads starts from various characteristics of roads,and uses various model methods such as graph theory,road Stroke,Mesh density,etc.,and has achieved good results,but there are also some shortcomings: for example,the consideration of the semantic features to roads is insufficient,at the same time,no contextual elements,such as residents in close contact with the road,were taken into account when selecting.In view of this,under the background to the rapid development of Big Data,this paper introduces POI(Point of Interest)data into the metrics of road semantic features,and comprehensively utilizes space syntax,road Stroke,and other model methods to consider the geometry,topology,distribution,semantic features of roads.A systematic study was made on the method of automatic selection for map generalization.The research achievements and innovations of the paper are as follow:(1)The POI data is introduced into the semantic analysis of road.On the one hand,people’s awareness of the importance to the road are related to the surrounding facilities around the road,on the other hand,the POI data are the representation form of the facilities around the road.The basic ideas are as follows: first,the POI data is reclassified,and then,the impact buffer of the road is determined according to the relevant standards,and then the three semantic feature parameters are calculated using the POI point falling into the buffer area.The semantic parameters are “the facility density”,the “important facility ratio” and the “thematic facilities ratio”.(2)A road automatic selection model based on road geometry,topology,distribution and semantic features is proposed.The basic idea is to calculate road geometry,topology and distribution characteristics based on road length,road width,connection value,control value and average line density,etc.Furthermore,calculate road semantic features combined with POI data,and then calculate the importance value of road.Finally,road stroke composition,stroke connectivity,etc.are selected as the constraints of road selection.Constraints are combined with road importance values to automatically select roads.In this paper,we use different types of actual road network data and POI data to carry out experiments on the proposed method.The experiment shows that introducing POI data into road semantic feature analysis can improve the selection probability of important road in people’s cognition and make the selection more reasonable.The proposed method of road selection combined with POI data can keep the main road after selection and keep the density difference in the distribution of road after selection,and also keep the connectivity of road.In the process of selection,the semantic feature information of the roads and facilities around the road is taken into account,and the selection probability of the important road in the people’s cognition is improved,and the scientific and rationality of the comprehensive results are improved.This research provides a new solution for the shortcomings of existing road automatic selection research.