Considering that the application of traditional Voronoi diagram in spatial division ignores the impact of road hierarchy,
speed, road-block, one-way and some other influential factors on proximity, an improved Voronoi diagram model based
on fuzzy interval theory is proposed in this paper by introducing different influential factors into the construction of
Voronoi diagram in order to enhance the accuracy of spatial division and to meet application requirements.
The idea of our improved Voronoi diagram model can be summarized as follows: Firstly, initial Voronoi diagram is built
via point by point insertion algorithm. Secondly, fuzzy interval on all Voronoi edges is generated via geometric
algorithm and is further used to represent each edge of Voronoi polygons. The size of fuzzy interval is determined by
taking all influential factors into consideration. Thirdly, an improved Voronoi diagram with fuzzy boundaries is provided
in which the proximity relationships between points in the fuzzy interval and the sites of Voronoi polygons which own a
public edge are all proposed to be the nearest. Finally, the validity and performance of our improved Voronoi diagram is
demonstrated through a typical application in emergency response system, in which the actual path distance is acted as
influential factor. Experimental results show that with our improved Voronoi diagram optimal attendance station can be
localized quickly and emergency response efficiency can be enhanced obviously. Besides, classification accuracy can be
increased more than 20% compared with traditional Voronoi diagram.