Search Results (144)

Individual tree crown detection (ITCD) and tree species classification are critical for forest inventory, species-specific monitoring, and ecological studies. However, accurately detecting tree crowns and identifying species in structurally complex forests with overlapping canopies remains challenging. This study was conducted in a complex mixed conifer–broadleaf forest in northern Japan, aiming to improve ITCD and species classification by employing two machine learning models and different combinations of metrics derived from very high-resolution (2.5 cm) UAV red–green–blue (RGB) and multispectral (MS) imagery. We first enhanced ITCD by integrating different combinations of metrics into multiresolution segmentation (MRS) and DeepForest (DF) models. ITCD accuracy was evaluated across dominant forest types and tree density classes. Next, nine tree species were classified using the ITCD outputs from both MRS and DF approaches, applying Random Forest and DF models, respectively. Incorporating structural, textural, and spectral metrics improved MRS-based ITCD, achieving F-scores of 0.44–0.58. The DF model, which used only structural and spectral metrics, achieved higher F-scores of 0.62–0.79. For species classification, the Random Forest model achieved a Kappa value of 0.81, while the DF model attained a higher Kappa value of 0.91. These findings demonstrate the effectiveness of integrating UAV-derived metrics and advanced modeling approaches for accurate ITCD and species classification in heterogeneous forest environments. The proposed methodology offers a scalable and cost-efficient solution for detailed forest monitoring and species-level assessment. Full article

Efficient and accurate acquisition of tree distribution and three-dimensional geometric information in forest scenes, along with three-dimensional reconstructions of entire forest environments, hold significant application value in precision forestry and forestry digital twins. However, due to complex vegetation structures, fine geometric details, and severe occlusions in forest environments, existing methods—whether vision-based or LiDAR-based—still face challenges such as high data acquisition costs, feature extraction difficulties, and limited reconstruction accuracy. This study focuses on reconstructing tree distribution and extracting key individual tree parameters, and it proposes a forest 3D reconstruction framework based on high-resolution remote sensing images. Firstly, an optimized Mask R-CNN model was employed to segment individual tree crowns and extract distribution information. Then, a Tree Parameter and Reconstruction Network (TPRN) was constructed to directly estimate key structural parameters (height, DBH etc.) from crown images and generate tree 3D models. Subsequently, the 3D forest scene could be reconstructed by combining the distribution information and tree 3D models. In addition, to address the data scarcity, a hybrid training strategy integrating virtual and real data was proposed for crown segmentation and individual tree parameter estimation. Experimental results demonstrated that the proposed method could reconstruct an entire forest scene within seconds while accurately preserving tree distribution and individual tree attributes. In two real-world plots, the tree counting accuracy exceeded 90%, with an average tree localization error under 0.2 m. The TPRN achieved parameter extraction accuracies of 92.7% and 96% for tree height, and 95.4% and 94.1% for DBH. Furthermore, the generated individual tree models achieved average Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) scores of 11.24 and 0.53, respectively, validating the quality of the reconstruction. This approach enables fast and effective large-scale forest scene reconstruction using only a single remote sensing image as input, demonstrating significant potential for applications in both dynamic forest resource monitoring and forestry-oriented digital twin systems. Full article

Accurate individual tree segmentation (ITS) in dense rubber plantations is a challenging task due to overlapping canopies, indistinct tree apexes, and intricate branch structures. To address these challenges, we propose a bottom-up, multi-feature fusion method for segmenting rubber trees using UAV-LiDAR point clouds. Our approach first involves performing a trunk extraction based on branch-point density variations and neighborhood directional features, which allows for the precise separation of trunks from overlapping canopies. Next, we introduce a multi-feature fusion strategy that replaces single-threshold constraints, integrating geometric, directional, and density attributes to classify core canopy points, boundary points, and overlapping regions. Disputed points are then iteratively assigned to adjacent trees based on neighborhood growth angle consistency, enhancing the robustness of the segmentation. Experiments conducted in rubber plantations with varying canopy closure (low, medium, and high) show accuracies of 0.97, 0.98, and 0.95. Additionally, the crown width and canopy projection area derived from the segmented individual tree point clouds are highly consistent with ground truth data, with R² values exceeding 0.98 and 0.97, respectively. The proposed method provides a reliable foundation for 3D tree modeling and biomass estimation in structurally complex plantations, advancing precision forestry and ecosystem assessment by overcoming the critical limitations of existing ITS approaches in high-closure tropical agroforests. Full article

Accurate detection and delineation of individual tree crowns (ITCs) are essential for sustainable forest management and ecosystem monitoring, providing key biophysical attributes at the individual tree level. However, the complex structure of mixed-wood forests, characterized by overlapping canopies of various shapes and sizes, presents significant challenges, often compromising accuracy. This study presents a two-stage deep learning framework that integrates Canopy Height Model (CHM)-based treetop detection with three-dimensional (3D) ITC delineation using high-resolution airborne LiDAR point cloud data. In the first stage, Mask R-CNN detects treetops from the CHM, providing precise initial localizations of individual trees. In the second stage, a 3D U-Net architecture clusters LiDAR points to delineate ITC boundaries in 3D space. Evaluated against manually delineated reference data, our approach outperforms established methods, including Mask R-CNN alone and the lidR itcSegment algorithm, achieving mean intersection-over-union (mIoU) scores of 0.82 for coniferous plots, 0.81 for mixed-wood plots, and 0.79 for deciduous plots. This study demonstrates the great potential of the two-stage deep learning approach as a robust solution for 3D ITC delineation in mixed-wood forests. Full article

Vertical structure monitoring of urban vegetation provides data support for urban green space planning and ecological management, playing a significant role in promoting sustainable urban ecological development. Three-dimensional green volume (3DGV) is a comprehensive index used to characterize the ecological benefit of urban vegetation. As a critical component of urban vegetation, street trees play a key role in urban ecological benefits evaluation, and the quantitative estimation of their 3DGV serves as the foundation for this assessment. However, current methods for measuring 3DGV based on point cloud data often suffer from issues of overestimation or underestimation. To improve the accuracy of the 3DGV for urban street trees, this study proposed a novel approach that used convex hull coupling k-means clustering convex hulls. A new method based on terrestrial laser scanning (TLS) data was proposed, referred to as the Convex Hull Coupling Method (CHCM). This method divides the tree crown into two parts in the vertical direction according to the point cloud density, which better adapts to the lower density of the upper layer of TLS data and obtains a more accurate 3DGV of individual trees. To validate the effectiveness of the CHCM method, 30 sycamore (Platanus × acerifolia (Aiton) Willd.) plants were used as research objects. We used the CHCM and five traditional 3DGV calculation methods (frustum method, convex hull method, k-means clustering convex hulls, alpha-shape algorithm, and voxel-based method) to calculate the 3DGV of individual trees. Additionally, the 3DGV was predicted and analyzed using five fitting models. The results show the following: (1) Compared with the traditional methods, the CHCM improves the estimation accuracy of the 3DGV of individual trees and shows a high consistency in the data verification, which indicates that the CHCM method is stable and reliable, and (2) the fitting results R² of the five models were all above 0.75, with the exponential function model showing the best fitting accuracy (R² = 0.89, RMSE = 74.85 m³). These results indicate that for TLS data, the CHCM can achieve more accurate 3DGV estimates for individual trees, outperforming traditional methods in both applicability and accuracy. The research results not only offer a novel technical approach for 3DGV calculation using TLS data but also establish a reliable quantitative foundation for the scientific assessment of the ecological benefits of urban street trees and green space planning. Full article

The accurate point cloud completion of individual tree crowns is critical for quantifying crown complexity and advancing precision forestry, yet it remains challenging in dense plantations due to canopy occlusion and LiDAR limitations. In this study, we extended the scope of conventional point cloud completion techniques to artificial planted forests by introducing a novel approach called Multi−feature Fusion Completion of Populus (MFCPopulus). Specifically designed for Populus Tomentosa plantations with uniform spacing, this method utilized a dataset of 1050 manually segmented trees with expert−validated trunk−canopy separation. Key innovations include the following: (1) a hierarchical adversarial framework that integrates multi−scale feature extraction (via Farthest Point Sampling at varying rates) and biologically informed normalization to address trunk−canopy density disparities; (2) a structural characteristics split−collocation (SCS−SCC) strategy that prioritizes crown reconstruction through adaptive sampling ratios, achieving a 94.5% canopy coverage in outputs; (3) a cross−layer feature integration enabling the simultaneous recovery of global contours and a fine−grained branch topology. Compared to state−of−the−art methods, MFCPopulus reduced the Chamfer distance variance by 23% and structural complexity discrepancies (ΔD_b) by 33% (mean, 0.12), while preserving species−specific morphological patterns. Octree analysis demonstrated an 89−94% spatial alignment with ground truth across height ratios (HR = 1.25−5.0). Although initially developed for artificial planted forests, the framework generalizes well to diverse species, accurately reconstructing 3D crown structures for both broadleaf (Fagus sylvatica, Acer campestre) and coniferous species (Pinus sylvestris) across public datasets, providing a precise and generalizable solution for cross−species trees’ phenotypic studies. Full article

Pre-trained foundation models, trained on large-scale datasets, have demonstrated significant success in a variety of downstream vision tasks. Parameter-efficient fine-tuning (PEFT) methods aim to adapt these foundation models to new domains by updating only a small subset of parameters, thereby reducing computational overhead. However, the effectiveness of these PEFT methods, especially in the context of forestry remote sensing—specifically for individual tree detection—remains largely unexplored. In this work, we present a simple and efficient PEFT approach designed to transfer pre-trained transformer models to the specific tasks of tree crown detection and species classification in unmanned aerial vehicle (UAV) imagery. To address the challenge of mitigating the influence of irrelevant ground targets in UAV imagery, we propose an Adaptive Salient Channel Selection (ASCS) method, which can be simply integrated into each transformer block during fine-tuning. In the proposed ASCS, task-specific channels are adaptively selected based on class-wise importance scores, where the channels most relevant to the target class are highlighted. In addition, a simple bias term is introduced to facilitate the learning of task-specific knowledge, enhancing the adaptation of the pre-trained model to the target tasks. The experimental results demonstrate that the proposed ASCS fine-tuning method, which utilizes a small number of task-specific learnable parameters, significantly outperforms the latest YOLO detection framework and surpasses the state-of-the-art PEFT method in tree detection and classification tasks. These findings demonstrate that the proposed ASCS is an effective PEFT method, capable of adapting the pre-trained model’s capabilities for tree crown detection and species classification using UAV imagery. Full article

Tropical forests are essential ecosystems recognized for their carbon sequestration and biodiversity benefits. As the world undergoes a simultaneous data revolution and climate crisis, accurate data on the world’s forests are increasingly important. Completely novel in approach, this study proposes a methodology encompassing two bespoke deep learning models: (1) a single encoder, double decoder (SEDD) model to generate a species segmentation map, regularized by a distance map in training, and (2) an XGBoost model that estimates the diameter at breast height (DBH) based on tree species and crown measurements. These models operate sequentially: RGB images from the ReforesTree dataset undergo preprocessing before species identification, followed by tree crown detection using a fine-tuned DeepForest model. Post-processing applies the XGBoost model and custom allometric equations alongside standard carbon accounting formulas to generate final sequestration estimates. Unlike previous approaches that treat individual tree identification as an isolated task, this study directly integrates species-level identification into carbon accounting. Moreover, unlike traditional carbon estimation methods that rely on regional estimations via satellite imagery, this study leverages high-resolution, drone-captured RGB imagery, offering improved accuracy without sacrificing accessibility for resource-constrained regions. The model correctly identifies 67% of trees in the dataset, with accuracy rising to 84% for the two most common species. In terms of carbon accounting, this study achieves a relative error of just 2% compared to ground-truth carbon sequestration potential across the test set. Full article

Structural changes in savanna trees vary spatially and temporally because of both biotic and abiotic drivers, as well as the complex interactions between them. Given this complexity, it is essential to monitor and quantify woody structural changes in savannas efficiently. We implemented a non-destructive approach based on Terrestrial Laser Scanning (TLS) and Quantitative Structure Models (QSMs) that offers the unique advantage of investigating changes in complex tree parameters, such as volume and branch length parameters that have not been previously reported for savanna trees. Leaf-off multi-scan TLS point clouds were acquired during the dry season, using a Riegl VZ1000 TLS, in September 2015 and October 2019 at the Skukuza flux tower in Kruger National Park, South Africa. These three-dimensional (3D) data covered an area of 15.2 ha with an average point density of 4270 points/m² (0.015°) and 1600 points/m² (0.025°) for the 2015 and 2019 clouds, respectively. Individual tree segmentation was applied on the two clouds using the comparative shortest-path algorithm in LiDAR 360(v5.4) software. We reconstructed optimized QSMs and assessed tree structural parameters such as Diameter at Breast Height (DBH), tree height, crown area, volume, and branch length at individual tree level. The DBH, tree height, crown area, and trunk volume showed significant positive correlations (R² > 0.80) between scanning periods regardless of the difference in the number of points of the matched trees. The opposite was observed for total and branch volume, total number of branches, and 1st-order branch length. As the difference in the point densities increased, the difference in the computed parameters also increased (R² < 0.63) for a high relative difference. A total of 45% of the trees present in 2015 were identified in 2019 as damaged/felled (75 trees), and the volume lost was estimated to be 83.4 m³. The results of our study showed that volume reconstruction algorithms such as TreeQSMs and high-resolution TLS datasets can be used successfully to quantify changes in the structure of savanna trees. The results of this study are key in understanding savanna ecology given its complex and dynamic nature and accurately quantifying the gains and losses that could arise from fire, drought, herbivory, and other abiotic and biotic disturbances. Full article

The accurate extraction of individual tree positions is key to forest structure quantification, and Unmanned Aerial Vehicle (UAV) visible light data have become the primary data source for extracting individual tree locations. Compared to deep learning methods, classical detection methods require lower computational resources and have stronger interpretability and applicability. However, in closed-canopy forests, challenges such as crown overlap and uneven light distribution hinder extraction accuracy. To address this, the study improves the existing Revised Local Maxima (RLM) method and proposes a Multi-Source Local Maxima (MSLM) method, based on UAV visible light data, which integrates Canopy Height Models (CHMs) and Digital Orthophoto Mosaics (DOMs). Both the MSLM and RLM methods were used to extract individual tree positions from three different types of closed-canopy stands, and the extraction results of the two methods were compared. The results show that the MSLM method outperforms the RLM in terms of Accuracy Rate (85.59%), Overall Accuracy (99.09%), and F1 score (85.21%), with stable performance across different forest stand types. This demonstrates that the MSLM method can effectively overcome the challenges posed by closed-canopy stands, significantly improving extraction precision. These findings provide a cost-effective and efficient approach for forest resource monitoring and offer valuable insights for forest structure optimization and management. Full article

Individual tree segmentation is crucial to extract forest structural parameters, which is vital for forest resource management and ecological monitoring. Airborne LiDAR (ALS), with its ability to rapidly and accurately acquire three-dimensional forest structural information, has become an essential tool for large-scale forest monitoring. However, accurately locating individual trees and mapping canopy boundaries continues to be hindered by the overlapping nature of the tree canopies, especially in dense forests. To address these issues, this study introduces CCD-YOLO, a novel deep learning-based network for individual tree segmentation from the ALS point cloud. The proposed approach introduces key architectural enhancements to the YOLO framework, including (1) the integration of cross residual transformer network extended (CReToNeXt) backbone for feature extraction and multi-scale feature fusion, (2) the application of the convolutional block attention module (CBAM) to emphasize tree crown features while suppressing noise, and (3) a dynamic head for adaptive multi-layer feature fusion, enhancing boundary delineation accuracy. The proposed network was trained using a newly generated individual tree segmentation (ITS) dataset collected from a dense forest. A comprehensive evaluation of the experimental results was conducted across varying forest densities, encompassing a variety of both internal and external consistency assessments. The model outperforms the commonly used watershed algorithm and commercial LiDAR 360 software, achieving the highest indices (precision, F1, and recall) in both tree crown detection and boundary segmentation stages. This study highlights the potential of CCD-YOLO as an efficient and scalable solution for addressing the critical challenges of accuracy segmentation in complex forests. In the future, we will focus on enhancing the model’s performance and application. Full article

Technological developments have allowed helicopter airborne laser scanning (HALS) to produce high-density point clouds below the forest canopy. We present a tree stem classification method that combines linear shape detection and model-based clustering, using four discrete methods to estimate stem diameter. Stem horizontal size was estimated every 25 cm below the living crown, and a cubic spline was used to estimate where there were gaps. Individual stem diameter at breast height (DBH) was estimated for 77% of field-measured trees. The root mean square error (RMSE) of DBH estimates was 7–12 cm using stem circle fitting. Adapting the approach to use an existing stem taper model reduced the RMSE of estimates (<1 cm). In contrast, estimates that were produced from a previously existing DBH estimation method (PREV) could be achieved for 100% of stems (DBH RMSE 6 cm), but only after location-specific error was corrected. The stem classification method required comparatively little development of statistical models to provide estimates, which ultimately had a similar level of accuracy (RMSE < 1 cm) to PREV. HALS datasets can measure broad-scale forest plantations and reduce field efforts and should be considered an important tool for aiding in inventory creation and decision-making within forest management. Full article

Natural secondary forests play a crucial role in global ecological security, climate change mitigation, and biodiversity conservation. However, accurately delineating individual tree crowns and identifying tree species in dense natural secondary forests remains a challenge. This study combines deep learning with traditional image segmentation methods to improve individual tree crown detection and species classification. The approach utilizes hyperspectral, unmanned aerial vehicle laser scanning data, and ground survey data from Maoershan Forest Farm in Heilongjiang Province, China. The study consists of two main processes: (1) combining semantic segmentation algorithms (U-Net and Deeplab V3 Plus) with watershed transform (WTS) for tree crown detection (U-WTS and D-WTS algorithms); (2) resampling the original images to different pixel densities (16 × 16, 32 × 32, and 64 × 64 pixels) and inputting them into five 3D-CNN models (ResNet10, ResNet18, ResNet34, ResNet50, VGG16). For tree species classification, the MSFB combined with the CNN models were used. The results show that the U-WTS algorithm achieved a recall of 0.809, precision of 0.885, and an F-score of 0.845. ResNet18 with a pixel density of 64 × 64 pixels achieved the highest overall accuracy (OA) of 0.916, an improvement of 0.049 over the original images. After incorporating MSFB, the OA improved by approximately 0.04 across all models, with only a 6% increase in model parameters. Notably, the floating-point operations (FLOPs) of ResNet18 + MSFB were only one-eighth of those of ResNet18 with 64 × 64 pixels, while achieving similar accuracy (OA: 0.912 vs. 0.916). This framework offers a scalable solution for large-scale tree species distribution mapping and forest resource inventories. Full article

Polarimetric measurement has been proven to be of great importance in various applications, including remote sensing in agriculture and forest. Polarimetric full waveform LiDAR is a relatively new yet valuable active remote sensing tool. This instrument offers the full waveform data and polarimetric information simultaneously. Current studies have primarily used commercial non-polarimetric LiDAR for tree species classification, either at the dominant species level or at the individual tree level. Many classification approaches combine multiple features, such as tree height, stand width, and crown shape, without utilizing polarimetric information. In this work, a customized Multiwavelength Airborne Polarimetric LiDAR (MAPL) system was developed for field tree measurements. The MAPL is a unique system with unparalleled capabilities in vegetation remote sensing. It features four receiving channels at dual wavelengths and dual polarization: near infrared (NIR) co-polarization, NIR cross-polarization, green (GN) co-polarization, and GN cross-polarization, respectively. Data were collected from several tree species, including coniferous trees (blue spruce, ponderosa pine, and Austrian pine) and deciduous trees (ash and maple). The goal was to improve the target identification ability and detection accuracy. A machine learning (ML) approach, specifically a decision tree, was developed to classify tree species based on the peak reflectance values of the MAPL waveforms. The results indicate a re-substitution error of 3.23% and a k-fold loss error of 5.03% for the 2106 tree samples used in this study. The decision tree method proved to be both accurate and effective, and the classification of new observation data can be performed using the previously trained decision tree, as suggested by both error values. Future research will focus on incorporating additional LiDAR data features, exploring more advanced ML methods, and expanding to other vegetation classification applications. Furthermore, the MAPL data can be fused with data from other sensors to provide augmented reality applications, such as Simultaneous Localization and Mapping (SLAM) and Bird’s Eye View (BEV). Its polarimetric capability will enable target characterization beyond shape and distance. Full article

This study compared hierarchical Bayesian, mixed-effects Gaussian process regression, and random forest models for predicting height to crown base (HCB) in Qinghai spruce (Picea crassifolia Kom.) forests using LiDAR-derived data. Both modeling approaches were applied to a dataset of 510 trees from 16 plots in northern China. The models incorporated tree-level variables (height, diameter at breast height, crown projection area) and plot-level spatial competition indices. Model performance was evaluated using leave-one-plot-out cross-validation. The Gaussian mixed-effects process model (with an RMSE of 1.59 and MAE of 1.25) slightly outperformed the hierarchical Bayesian model and the random forest model. Both models identified LiDAR-derived tree height, DBH, and LiDAR-derived crown projection area as primary factors influencing HCB. The spatial competition index (SCI) emerged as the most effective random effect, with the lowest AIC and BIC values, highlighting the importance of local competition dynamics in HCB formation. Uncertainty analysis revealed consistent patterns across the predicted values, with an average relative uncertainty of 33.89% for the Gaussian process model. These findings provide valuable insights for forest management and suggest that incorporating spatial competition indices can enhance HCB predictions. Full article