When the traditional Voronoi diagram divides a large number of point sets into Voronoi, there are too many Voronoi cells, which makes it difficult to apply to such fields as geographic information systems and biomedicine. In order to solve this problem, a Voronoi diagram based on adaptive DBSCAN (density-based spatial clustering of applications with noise) was proposed. The phenomenon of Voronoi unit merge was explained. The necessary and sufficient conditions for its occurrence were proved. The algorithm for generating the Voronoi diagram was proposed and simulated. In order to verify its effectiveness, the algorithm was applied to neutrophils under the microscope and fire point data on the surface of China. The results show that the algorithm can effectively solve the problem that Voronoi diagram is too meticulous when the point set size is large, which breaks through the single point to single point division form of the traditional Voronoi diagram. In addition,the algorithm broadens the application of Voronoi diagrams in the fields of graphic image processing,biomedicine and geographic information systems.