Based on airborne LiDAR data for individual tree segmentation,this article discusses the individual tree extraction method based on canopy height model(FCHM)and the individual tree extraction algorithm based on normalized point cloud(FNPC).The root mean square value of the total accuracy of FCHM's segmentation of individual trees is 0.93,the root mean square value of F_Score is 0.62,and the total accuracy and F_Score of FNPC algorithm is 0.90 and 0.56,respectively.The accuracy of FCHM algo-rithm is better than that of FNPC algorithm.Through analysis of different forest forms,forest types,and stand densities,it was found that the extraction effect is best in a single layer coniferous forest with a stand density of less than 400 trees/hm2.
关键词
单木分割/LiDAR点云/冠层高度模型/归一化点云/NEWFOR公开数据集
Key words
individual tree segmentation/LiDAR point cloud/the canopy height model/normalized point cloud/NEWFOR public dataset
维普
万方数据