DefinePK

Abstract

As an important method of spatial analysis, road network density
analysis is widely used in intelligent transportation, urban planning and automatic
map generalization. Classical grid-based method for solving road network density
is widely adopted in GIS. However, this method has the disadvantages of low
accuracy and lack of general applicability. Through a survey of the road network
density analysis method, this work proposes a novel optimized analysis method
based on skeleton partitioning with constrained Delaunay Triangulated Irregular
Network D-TIN. By means of breaking up the whole into parts, the method calculates
the area contributions of all associated triangulations to the road growth region
and in this way the road network density was obtained finally. It simplifies
the process of road network density solving. The experiment shows that the optimized
D-TIN method is robust and efficient in the whole calculation process. In the
real application scenarios of 150000-1100000 comprehensive road network map
automatic selection and generalization, our test indicates that the optimized
D-TIN method can provide an efficient means of road network automatic map generalization
and can be applied to the quality evaluation of the road network generalization.