Extraction of branch factors and model construction for Larix plantation using terrestrial laser scanning (TLS)

Forest parameters could be estimated without damage using terrestrial laser scanning (TLS). In this study, 1266 sets of branch parameters, including branch height, branch chord length, branch length, branch angle, branch diameter and the branch arc height, of 26 Larix trees were extracted based on TLS using point cloud processing software in the way of human-computer interaction. The average of extracted maximum relative height of the branches was 0.83. The accuracy of the extracted branch attributes followed the order of branch height > branch chord length > branch length > branchdiameter (with the branch diameter >20 mm) > branch arc height. After dividing the crown into four parts, it proved that the density of the branches increased while the extraction ratio and precision of branches decreased with the increase of crown height. Due to the low accuracy of branch diameter extraction, we developed a base diameter prediction model using the branch chord length, the height of branch, diameter at breast height and tree height as independent variables. The measured, extracted and model predicted values of different base diameter were compared, and discovered that accuracy of model prediction was greater than that of extraction by TLS. For timber production, the most worthy wood was the middle and lower section of the tree. The method could accurately extract tree height, diameter at breast height (DBH) and the branch attributes with the relative branch height of less than 0.8, providing the parameters that are necessary to develop a wood quality model.