Single Tree Segmentation and Diameter at Breast Height Estimation With Mobile LiDAR

The tree diameter at breast height (DBH) is one of the most important variables for monitoring forest ecology. Mobile laser scanning (MLS), which has been widely applied in forestry field, makes DBH measurement fast and convenient. However, there are many shrubs deadwood neutral environment point clouds quality from MLS easily affected by results low single segmentation estimation accuracy. To improve accuracy a complex cloud quality, we propose relative density method based on multi-height diameters estimation. calculates according to ratio two different scales, segments trunks higher trunk compared with their surroundings points. In natural plot, precision recall proposed reached 0.966 0.946, respectively; urban 1 0.966, respectively. was used estimate trees using diameters. combined outlier detection algorithm were able robustness when have large noise. For results, mean absolute error (MAE), percentage (MAPE), root square (RMSE) 2.5 cm, 11.54%, 3.17 respectively, plot 1.65 6.31%, 1.97 plot. good experiment indicate that can achieve accurate robust extraction provide fundamental data supervision sustainable development resources.