Answers are grounded in this paper and cite their sources.
Comparative study reveals marked accuracy drops in deep learning tree segmentation when tested against laser scanning in closed-canopy forests, indicating manual labelling inflates performance.
Monitoring forest dynamics at an individual tree scale is essential for accurately assessing ecosystem responses to climate change, yet traditional methods relying on field-based forest inventories are labor-intensive and limited in spatial coverage. Advances in remote sensing using drone-acquired RGB imagery combined with deep learning models have promised precise individual tree crown (ITC) segmentation; however, existing methods are frequently validated against human-annotated images, lacking rigorous independent ground truth. In this study, we generate high-fidelity validation labels from co-located Terrestrial Laser Scanning (TLS) data for drone imagery of mixed unmanaged boreal and Mediterranean forests. We evaluate the performance of bounding box predictions from two widely used deep learning ITC segmentation models - DeepForest (RetinaNet) and Detectree2 (Mask R-CNN) - on these data, and compare to performance on further Mediterranean forest data labelled manually. When validated against TLS-derived ground truth from Mediterranean forests, model performance decreased significantly compared to assessment based on hand-labelled from an ecologically similar site (for all trees with crowns visible from above AP50: 0.117 vs. 0.670). Restricting evaluation to only canopy trees (here defined as trees attaining a height of at least 75\% of the maximum plot height) shrank this gap considerably (Canopy AP50: 0.379), although performance was still far lower than on similar hand-labelled data. Models also performed poorly on boreal forest data (all trees AP50: 0.176), although again increasing when evaluated on canopy trees only (AP50: 0.363). Compared to manual labels, performance was particularly weak for recall of TLS labels, while precision was closer. Both models showed very poor localisation accuracy (that is, alignment of prediction with TLS label) at stricter IoU thresholds, even when restricted to canopy trees (Max AP75: 0.047). Similar results have been observed in studies using aerial LiDAR data, suggesting fundamental limitations in aerial-based segmentation approaches in closed canopy forests.
No takes yet. Share an insight, caveat, or question.
Allen et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: