An innovative and automated method for characterizing wood defects on trunk surfaces using high-density 3D terrestrial LiDAR data

We designed a novel method allowing to automatically detect and measure defects on the surface of trunks including branches, branch scars, epicormics from terrestrial LiDAR data by using only high-density 3D information. could with diameter as small 0.5 cm either oak (Quercus petraea (Matt.) Liebl.) or beech (Fagus sylvatica L.) trees that display rough smooth bark. Ground-based light detection ranging (LiDAR) technology describes standing high level detail. This provides an opportunity assess tree quality use this information in forest inventory. Assuming availability very detail, we extract about defects, mainly inherited past ramification having strong impact wood quality. Within general framework development computing able detect, identify, quantify trunk described produced LiDAR, study focuses relevance whole process for two species contrasted bark roughness Liebl. Fagus terms detection, identification comparison measurements performed manually surface. First, segmentation algorithm detected singularities Next, Random Forests machine learning identified most probable defect type allowed elimination false detections. Finally, estimated position, horizontal, vertical dimensions each data, compared them those observed directly operator. The were classified accuracy average $${F}_{1}$$ score (harmonic mean precision recall) 0.74. There differences computed areas, but much closer agreement number defects. present measured can be used automated procedure grading roundwoods.