Abstract: UAV-borne LiDAR point cloud data is an important data source for producing DEM. In order to further improve DEM production efficiency,selecting flat terrain and mountainous terrain as test areas,the ground point cloud,which is processed by filtering method,is thinned and simplified according to the algorithm based on TIN with seven different the ground point cloud retention rate of 80%,60%,40%,and so on,and the corresponding DEM is generated and its accuracy is evaluated by mean error (ME),standard deviation (SD),and root mean square error (RMSE). The results show that: ①The accuracy of the produced 0.5 m grid-spacing DEM could exceed 0.05 m when the ground point cloud density reached 2 points/m² for flat terrain and 9 points/m² for mountainous terrain. ②As the density of ground point cloud increases,the DEM accuracy level gradually stabilizes,and the DEM accuracy would decrease rapidly when the ground point cloud density is thinned to 1 point/m². For the DEM production tasks in large regions using UAV-borne LiDAR point cloud data,the conclusions of this research have a certain guiding and reference significance in reducing data acquisition costs and improving DEM production efficiency.

Key words: point cloud density, digital elevation model, LiDAR point cloud, point cloud thinning, UAV-borne