In order to analyze the similarities and differences between the UAV dense matching point cloud and the airborne LiDAR point cloud,the spatial distribution of two kinds of point clouds for dense forest(canopy density of 0.85),sparse forest(canopy density of 0.55) and undeveloped forest was visually analyzed.The densely matched point cloud was normalize through two types of DEM(UAV_DEM and LiDAR_DEM) produced by UAV and LiDAR point clouds respectively.Then,the statistical characteristics of the obtained two sets of normalized densely matched point cloud data and LiDAR point cloud data were compared using the paired sample t-test analysis.The results show that:1) In dense forests,densely matched point cloud has great limitations to obtain canopy internal and ground information.After normalization using LiDAR_DEM,the densely matched point cloud and the laser point cloud were examined with significant differences in the middle and lower quantile heights and all density characteristics(α=0.05),while no significant differences found in the middle and upper quantile heights;2) In sparse forest and undeveloped forest,except for the lower quantile height,the remaining statistical parameters of height and density are not significantly different between the densely matched point cloud and the laser point cloud.However,the densely matched point cloud is superior to the airborne LiDAR point cloud regarding the ability of describing the three-dimensional structure of young trees.In conclusion,the UVA dense matching point cloud can be directly used to estimate the forest parameters of sparse forest and undeveloped forest in the forest survey and monitoring.With the assistance of high-precision DEM,some parameters of dense forest(e.g.,crown layer height) can be further estimated through the UVA dense matching point cloud.