Extension and application of arc segment data structure
Wang Jie-Chen,Li Yong-Quan,Qian Chen-Hui
DOI: https://doi.org/10.3321/j.issn:0469-5097.2008.01.008
2008-01-01
Abstract:The spatial data structure is one of the most elementary problems in geographic information systems(GIS).It is also the foundation GIS depends on to organize,store,analyze and display spatial information,and arc is one most important element for vector data structure.The existing vectorial arc data structure can store and organize coordinate information effectively,but some vector-based calculations are inefficient.Because there are often large numbers of nodes in one arc,as some kinds of spatial indexing for vector-based operations,the extent of the whole arc is clearly very rough.Enlightened by the idea of spatial index,dynamical partition,exchanging space for time and so on,the paper proposed to handle the whole arc with "logical partition",obtain spatial limit information of each segment and build up to arc data structure.As for each subsection,we restrict the total number of nodes strictly,calculate and store its coordinate extent beforehand by adding information of arc data structures,which are NumParts(number of logical segments) and PartExtents(Extent of the each logical segment).In this paper,we divide arc into many parts with identical node number n.In other words,the numbers of node in each part are n.The paper also gives realizing method for this modified structure and analyses time efficiency and space efficiency in some frequently used procedures.Take the computation of the distance between point and arc for example,the paper analyses the relationship between computational efficiency and sub-arc node number and then gives practical results using artificial and real arcs from a GIS environment.The tests include: Through the distance calculation from fixed point to different arc segments in conditions of different partition,we test the effect on computational efficiency caused by the number of points in test arc segments and the number of points in each subsection,and compare it with the situation without partition;Through the distance calculation from different points to the same arc segment,we analyze the effect on computational efficiency caused by the distance from the test point to arc segment;And we test the average capacity in rather big bound.Test results indicated that,the algorithm requires a little more additional memory cost than the majority of other approaches,but we are rewarded with evidently higher computational efficiency if the extended data structure is employed.In cases of massive data environment,real-time response process,and some spatial analysis problems that need to calculate distance from point to arc frequently,the method of this paper can improve efficiency obviously and shorten system response time.